AustLII Home | Databases | WorldLII | Search | Feedback

University of New South Wales Law Journal

Faculty of Law, UNSW
You are here:  AustLII >> Databases >> University of New South Wales Law Journal >> 1999 >> [1999] UNSWLawJl 43

Database Search | Name Search | Recent Articles | Noteup | LawCite | Author Info | Download | Help

Ghosh, Eric --- "Applying Pettit's Republican Liberty to Criminal Justice and Judicial Decision-Making: The Need for Other Values Including Desert and a Suggestion that they be Understood Consequentially" [1999] UNSWLawJl 43; (1999) 22(1) UNSW Law Journal 122

APPLYING PETTIT'S REPUBLICAN LIBERTY TO CRIMINAL
JUSTICE AND JUDICIAL DECISION-MAKING: THE NEED FOR
OTHER VALUES INCLUDING DESERT AND A SUGGESTION
THAT THEY BE UNDERSTOOD CONSEQUENTIALLY

ERIC GHOSH[*]

I. INTRODUCTION

In the United States some esteemed constitutional scholars, including Frank Michelman[1] and Cass Sunstein,[2] have argued that republican political theory should inform legal reasoning. They have relied upon historians who have suggested that republican thought had been influential in America in the eighteenth century, when America obtained its independence and adopted its constitution. This resurgence of interest in republican theory, together with the question of whether Australia should become a republic,[3] has sparked interest here in the relationship between republican theory and Australian law. This relationship can be considered in terms of the following two issues. First, does the limited authority of judges and other agents of our political system, such as politicians, bureaucrats and police officers, constrain them from applying republican theory to their work? It might be argued that judges, for instance, can only rely upon moral theories which hold a sufficient place within the country's legal and political history. There is controversy over whether republican political theory meets this criterion.[4] The second issue is whether republican theory should guide agents who are not constrained from using it. This depends upon whether republican theory is attractive.

The Australian philosopher, Philip Pettit, has been concerned with this second issue: he has attempted to present an attractive understanding of republican theory centred on the value of liberty.[5] He has also applied republican liberty to the criminal justice system in a joint project with the Australian criminologist John Braithwaite, entitled Not Just Deserts.[6] They argue in favour of a less punitive criminal justice system, a system guided at all levels by republican liberty.

The question of whether judges should be guided by republican liberty has been raised more recently by Braithwaite's republican model of judicial deliberation.[7] Braithwaite suggests that judges could apply Pettit's republican liberty in hard cases or at least in hard cases relating to criminal justice.[8] Most recently, Pettit has published a book, Republicanism, in which he elaborates upon his understanding of republican liberty and outlines its broad political implications.[9] This makes a significant contribution to republican theory and provides additional justification for taking Pettit's interpretation seriously. Nevertheless, Braithwaite and Pettit's book on criminal justice still provides the most thorough discussion of republican liberty's implications for a single area of law. While this book covers a wide range of issues concerning the criminal justice system, I focus on their justification for punishment.[10] I evaluate this by considering whether republican liberty furnishes a sufficiently strong basis for upholding certain rights that protect people from being punished for offences they have not committed.

First, I indicate how Pettit understands republican liberty and its implications. Where Pettit's ideas have changed, I draw on his more recent book. I then consider some differences between deontology and consequentialism in order to indicate why Pettit's approach could represent a significant advance in moral theory if it succeeds. While deontology is often regarded as providing stronger support for rights, Pettit nevertheless places republican liberty within a consequentialist framework. I then refer to some traditional theories of punishment to indicate why Braithwaite and Pettit feel it necessary to propose a new approach. In Part III, Pettit's theory is tested by applying it to a hypothetical case concerning a detective who is considering whether to manufacture evidence.[11] This hypothetical suggests that Pettit's theory provides insufficient reasons for why agents of the criminal justice system, including detectives and judges, should respect uncontroversial rights. Part IV considers whether Pettit's theory provides sufficiently powerful arguments in favour of the qualified, but nevertheless controversial, 'right' of an accused person to be provided with legal representation, that the High Court found in Dietrich v R.[12]

On this point, Pettit's approach compares unfavourably with Michelman's republican model of judicial deliberation.[13] Michelman's model is itself controversial but I rely on it mostly to elaborate a conception of reasoning in which judges are open to the plurality of moral considerations relevant to legal issues. A major value missing from Pettit's theory is desert. I suggest that desert and other values could be incorporated within a consequentialist framework.

Not Just Deserts has received favourable[14] as well as critical comments. Referring to the latter, Sadurski examines the book's claim that having republican liberty as the single target for the criminal justice system avoids the pitfalls of retributivist and utilitarian theories and the complexity of mixing these approaches. He questions whether the target of republican liberty has these advantages, by demonstrating that republican liberty is in fact "a mixed bag of distinct concerns, each of them pointing sometimes in different directions".[15] I agree, but am nevertheless concerned with the opposite problem. Rather than emphasising the inclusive quality of republican liberty, I argue that it excludes too much. In a brief review, Kleinig notes that Braithwaite and Pettit never really show why mixed theories must be excessively complex.[16] I reinforce this point by arguing that their attempts to achieve simplicity come at too high a cost. Von Hirsch and Ashworth believe that Braithwaite and Pettit's adoption of consequentialism leads to the exclusion of desert and therefore to unjust trade-offs between individuals and the community.[17] While I agree that desert should not be excluded, accepting desert does not require the rejection of consequentialism.

II. REPUBLICAN LIBERTY

A. Some Characteristics of Republican Liberty

Pettit traces the republican tradition to Rome and points out that it was associated with the republics of Renaissance Italy (eg Machiavelli), seventeenth century England and eighteenth century France and America. What unites republican thought is a belief that a republic requires the rule of law, a system of checks and balances, and virtue (there must be a sufficient number of citizens prepared to demonstrate commitment to the public good). Support for these three features is motivated by a desire to check the exercise of arbitrary power so that liberty can flourish.[18]

Pettit refers to Berlin's influential discussion of liberty, in which Berlin argued that negative liberty involves the absence of interference while positive liberty involves agents taking an active part in gaining mastery of themselves: the self which they identify with must take charge of the lesser or partial selves that lurk within every individual.[19] Liberalism is traditionally associated with negative liberty while republican theory is often linked to positive liberty. For example, Michelman invokes republicanism to suggest that one is free as an individual only if one's choices are informed by critical reflection.[20] Furthermore, freedom requires participation in collective deliberation on the norms which should govern the community.[21] While Michelman does not explicitly relate participation in collective deliberation to mastery, there are several implicit links. Public discussion with others can prompt critical reflection by individuals on their own values and may enlarge their vision so that rather than being governed by a lower self which takes a narrow perspective, they identify with the good of the community. By directly participating in political decision-making, members of the community become masters of their own destiny rather than being passive recipients of government decrees. Republicans such as Michelman understand participation as an important element of liberty itself, while liberals value political participation because of its role in protecting citizens from state interference.[22]

However, Pettit argues that linking republicanism to positive liberty is a mistake which follows from liberals associating republicanism with continental Romantics such as Rousseau and Hegel.[23] Pettit suggests instead that republicans were concerned with a form of liberty which is neither wholly positive nor negative, but rather contains elements of each.[24] This is best understood by mentioning how Pettit describes liberal and republican understandings of the relationship between liberty and law. I am not concerned with evaluating his depiction of the political traditions of liberalism and republicanism but only with explaining how Pettit understands republican liberty. This allows for later consideration of whether Pettit has provided an attractive moral theory for the criminal justice system.[25]

Pettit suggests that when liberals in the eighteenth and nineteenth centuries started to think about liberty, they did so in the context of concern with state interference, particularly in trade and commerce.[26] For them, the antonym of liberty was any form of restraint or interference. Even law itself was an invasion of liberty, although it might be justified if it increased liberty overall by preventing greater assaults on liberty.[27] By contrast, the republican notion of liberty arose in the classical and medieval worlds with law constituting some people as citizens rather than as slaves. Here, the antonym of liberty was slavery or subjection. It was law which constituted one as a citizen rather than a slave, so law was seen as integral to liberty rather than an invasion of it. Republicans might see law as interference but liberty is not diminished by law unless that interference is arbitrary.[28]

Pettit also says that liberal liberty is only concerned with actual interference by others; it does not require the conditions which give people a sense of security.[29] However, under a republican conception, liberty requires not just a lack of interference but an assurance that the law actually protects you from such interference; only then can people feel free.[30] It is precisely this protection which slaves lack. Even if a slave has a benevolent, non-interfering master, the slave still has a master, that is someone who enjoys a significant, unchecked discretion to interfere.[31] This slave cannot enjoy the psychological status of being an equal although he or she may suffer less fear and may not have to be as deferential as a slave with a less benevolent master. Pettit says that republicans see the point of liberty as the achievement of a life where one feels secure and able to look others in the eye.[32] Republican liberty, then, extends beyond the liberal concern with actual non-interference, to the concern that people are legally protected from interference. On the other hand, it is only concerned with arbitrary interference while liberals are concerned about interference generally. These two differences mean that republican liberty involves an absence of mastery (or domination) by others. However, this is not the same as self-mastery (or positive liberty). At the individual level, a freed slave enjoys republican liberty but may, nevertheless, behave impulsively or habitually. At the collective level, political participation is defended by Pettit on the ground that it is instrumental in protecting against arbitrary interference by the state, rather than being justified on the ground that it is a defining core of liberty. Pettit's republicanism places less emphasis upon direct participation in politics.

Apart from Pettit's republican liberty involving aspects of both negative and positive liberty, two other details need to be mentioned. First, republican liberty comes in degrees of intensity and extent.[33] Intensity relates to the degree of arbitrary interference that the agent is exposed or vulnerable to. Extent refers to the choices available to the agent, due to factors other than arbitrary interference.

Thus, an obligation to make financial contributions imposed by a legitimate taxation system does not reduce intensity of freedom since the interference is not arbitrary, but it does reduce the extent of freedom since it leaves the agent with less money and, consequently, fewer options.[34] Pettit suggests that the state should give priority to intensity over extent of freedom; it is intuitively correct to reverse domination first and only then maximise the range and ease of choice.[35]

Secondly, maximising republican liberty requires an egalitarian distribution of that liberty.[36] Pettit suggests that an individual's intensity of freedom (his or her degree of protection from arbitrary interference) is a function of the individual's power relative to others. The more powerful one person becomes, the more vulnerable others become. Furthermore, enhancements in one's power to resist interference are subject to diminishing marginal productivity.[37] The same increment of increased power will achieve a greater increase in the level of protection enjoyed by a person on a low level of security than it would achieve for someone who already has a high level of security. These factors suggest that intensity of freedom is best optimised through its equal distribution.[38] He says that those who are unpersuaded of this may take equally intense freedom as society's goal.[39]

In summary: republican liberty has a psychological aspect (it is not only concerned with whether you happen to be free from interference but with whether you are in a position to feel secure); it is particularly concerned with arbitrary interference; and it suggests that freedom from arbitrary interference should be equally distributed. While this provides sufficient detail concerning the content of republican liberty, I finally note that Pettit places this value in a consequentialist framework. He claims that having republican liberty as our consequentialist target avoids the significant difficulties confronting deontological theories. The next sub-section describes the distinction between deontology and consequentialism, it mentions some alleged weaknesses of deontology and indicates two approaches to addressing a particular difficulty facing consequentialism. Both consequentialism and deontology have vigorous philosophical defenders and are based on intuitions which many people are deeply committed to. I think that consequentialism has advantages over deontology which suggest that Pettit's attempt to provide a consequentialist target deserves serious consideration. In the following sub-section, I merely indicate some of the advantages; I do not offer a sustained defence of consequentialism. I also refer to Scheffler's hybrid theory to indicate that a consequentialist target can co-exist with a permission for agents to act, up to a point, on deontological intuitions. Accepting deontological intuitions does not necessarily entail a rejection of attempts to define a consequentialist target. However, readers who reject consequentialism can interpret this article as suggesting that Pettit's approach has significant problems which are separate from its consequentialism.[40]

B. Consequentialism, Deontology and Criminology

Consequentialist theories state that we should do what produces the best consequences for society as a whole.[41] Deontological theories, on the other hand, require us to adhere to certain constraints, regardless of the consequences. They commit us to respecting certain rights even if their violation is necessary to stop others having identical rights violated. Thus, an agent is not allowed to kill one person even if this is the only way the murder of two others by someone else can be prevented. However, the consequentialist would say that two murders is a worse outcome than one murder, so the agent should kill that one individual. Consequentialists regard the deontological position as paradoxical. They suggest that deontological constraints are motivated by strong opposition to the violation of certain rights or constraints, yet deontologists favour here an outcome which maximises rather than minimises the violation of rights.[42] While this is best illustrated in the context of competing identical rights, in reality, different rights tend to be involved. However, the broad point remains that deontology constrains agents from achieving outcomes which involve the least violation of rights. One qualification to this is that non-absolutist deontologists accept that constraints may lose their binding quality in the face of disastrous consequences.[43] While there is some arbitrariness with respect to the point at which consequences become disastrous, it must involve more than two deaths (otherwise we have a consequentialist theory) and should involve significantly more if deontology is to appear distinct from consequentialism.

Deontologists also see difficulties in consequentialism.[44] One major difficulty is that consequentialism requires too much from agents. Can we really condemn an agent who refuses to kill a person in order to save two, on the basis of firmly held deontological intuitions? One consequentialist response is to state that while refraining from murder is incorrect here, the agent should not be condemned. Overall, we benefit from living in a society where people feel under strong moral constraints not to kill, even if this occasionally leads to sub-optimal results.[45] An alternative approach is to give agents a limited prerogative not to adopt an impersonal standpoint.[46] Scheffler argues that requiring agents to take an impersonal standpoint violates their integrity. A fundamental aspect of individuals is that they have personal projects and commitments that have moral weight which is independent of the weight those projects and commitments have in an impersonal calculus.[47] To require agents to only give such weight to their own projects as an impersonal calculus warrants is to violate a fundamental aspect of personhood that society values. Rather than alienating people from their particular projects and commitments, agents should be allowed to pursue them to a greater extent than an impersonal standpoint allows.[48] It is still morally correct for the agent to kill one person to save two, but a refusal to kill cannot be condemned because the agent has a prerogative here not to pursue what is consequentially correct. While deontologists simply prohibit agents from acting in a consequentially optimal manner, Scheffler's approach permits such action but does not condemn an agent who instead follows a deontological constraint. Rather than representing an unprincipled compromise between consequentialism and deontology, Scheffler grounds his concession to deontology on respect for people's integrity.[49]

These approaches avoid the deontological paradox of maximising the violation of rights. At the same time, agents are not always required to take an impersonal standpoint. However, the major consequentialist theory, utilitarianism, permits repugnant trade-offs between individual rights and general welfare. It can, in principle, favour convicting an innocent individual when the individual's disutility is outweighed by the satisfaction and security thereby obtained for the rest of the community.[50] Pettit argues that this problem of victimisation does not inevitably accompany consequentialism since it can be avoided by replacing utility with republican liberty. We then have a consequentialist theory which supports strict adherence to uncontroversial rights.[51] If he succeeds, this could represent a considerable advance in moral theory, assuming one agrees with Pettit that there are significant advantages in avoiding deontological theories. I have already mentioned the deontological paradox. However, Pettit is also concerned about complexity. Deontological approaches only determine constraints, so their adherents must subscribe to other moral principles to guide behaviour not determined by the constraints and these other principles are likely to be based on consequentialist targets.[52] This mix of constraints and targets contrasts with the simplicity of a consequentialist theory resting on the promotion of a single value. Pettit says that a normative theory must be complex enough to deal with the real world: "[w]hat simplicity requires is that it is not more complex than necessary".[53] Pettit argues that republican liberty bypasses deontological complexity while avoiding the problem of victimisation that utilitarianism encounters. I consider the issue of victimisation in the next Part.

Before turning to this, I should briefly place Braithwaite and Pettit's work in a criminological context.[54] Braithwaite and Pettit mention that until the 1970s, the dominant criminological theory was a species of utilitarianism which they call "preventionism".[55] The idea was that punishment should promote the goal of preventing crime through incapacitation (eg, imprisonment prevents inmates committing certain crimes), deterrence and rehabilitation. At a theoretical level, this approach seemed capable of justifying punishment of the innocent. At a practical level, there were less extreme difficulties which led to disillusionment with preventionism and to the revival of retributivism in the 1970s. Retributivists believe that criminals should be punished because they deserve it and only at a level which reflects their culpability. They argued that preventionism violated this principle; for example, rehabilitation was used to justify excessive sentences.[56] Furthermore, the idea that criminals need to be `treated' involved an assumption that criminals were determined creatures whose behaviour could not be accounted for by their own choices to break the law. Braithwaite and Pettit, however, are concerned that retributivist theories play into the hands of conservative law-and-order politicians who appeal to simple-minded vengeance,[57] and furthermore, that these deontological theories are plagued by a range of theoretical difficulties. They suggest that their approach avoids these difficulties and also the victimisation problem that crime preventionism encounters. I will now test whether republican liberty avoids the victimisation problem by considering a hypothetical case of a detective tempted' to manufacture evidence against an innocent person. While this may appear removed from judicial decision-making, it is useful in examining Pettit's arguments.

III. THE CASE OF THE DETECTIVE

A. Promoting Republican Liberty

Assume that a police detective, Mr Holmes, is charged with finding a sadistic serial killer who targeted children.[58] The murders took place in a small town and have terrified its inhabitants: the attacks occurred both in children's homes and in public spaces. Mr Holmes is fairly certain that the perpetrator was Mr Jones, who has died recently. Mr Holmes had some evidence which raised a suspicion that Mr Jones was involved, so he interviewed the dying Mr Jones and obtained an oral confession. Mr Jones was a highly respected and well-liked figure in the community. With the town now in mourning, an accusation against Mr Jones on the basis of an alleged oral confession and some evidence that raises no more than a suspicion would be greeted with outrage and disbelief. Mr Holmes also knows that the population will only feel secure once it is convinced that the perpetrator is dead or behind bars. He is considering, then, whether he should manufacture some evidence to secure a conviction against an individual who can be easily set up. Mr Smith is such an individual: he was released from jail before the serial attacks began, having served time for assaults on children. Utilitarianism could suggest that Mr Holmes should manufacture evidence against Mr Smith: Mr Smith is used to jail and may not be highly distressed to return there. On the other hand, the community's utility may increase considerably since it will feel more secure and will also feel satisfied that the crimes have been avenged.[59] Employing the theory of crime preventionism, the goal of deterrence may be better served through convicting Mr Smith than by convicting no one. Also, given his record, Mr Smith might commit further offences so incapacitating him could reduce crime.

Pettit would suggest that republican liberty be used instead of utility or crime prevention. In other words, Mr Holmes should only be concerned with maximising people's freedom from arbitrary interference and their sense of security. The community's freedom from arbitrary interference is perhaps enhanced slightly by imprisoning Mr Smith since there is a possibility that he might re-offend. On the other hand, the community's sense of security may be enhanced significantly by convicting him. While this suggests framing Mr Smith, Pettit would argue that this would be foolish since the fabrication of evidence might become known. First, it is likely that Mr Smith will know. He will tell others and this will at least raise doubts in their minds about the integrity of Mr Holmes. Furthermore, the entire community might become aware of the fabrication. The community will begin to question whether the true perpetrators of certain crimes are in fact behind bars and may also wonder whether they might themselves be framed by the police. Manufacturing the evidence has the potential to do such devastating damage to the community's sense of security that it cannot be contemplated as a measure to promote security even if the probability of the fabrication being exposed is small.[60] Security can only be promoted if Mr Holmes binds himself to respect certain rights.

Pettit would say that Mr Holmes must not only bind himself to respect certain rights; he must also assure the public that he does so. It may not be enough to simply tell the public that he respects rights, but instead, he may have to go to some pains to demonstrate his commitment.[61] We could point out that it is particularly in circumstances where there are strong pressures to violate rights that Mr Holmes can most convincingly affirm his commitment.[62] By manufacturing evidence in this situation, Mr Holmes misses an opportunity to affirm his respect for rights. The community will be aware of the pressures upon Mr Holmes and will obtain some reassurance from the fact that no one is charged or convicted in relation to the offences.

While these arguments have some cogency, they may not persuade Mr Holmes. He may reply that while Mr Smith could succeed in raising some doubts among his close associates about Mr Holmes' integrity, his associates are few in number and any loss of security by them would be outweighed by the gains in security obtained by the rest of the community believing that the killer is in jail. With respect to the community finding out that the evidence was manufactured, Mr Holmes might say that there is simply no way that the falsity of some of the evidence could be proved. Furthermore, even if the falsity was exposed, the public will not become exceedingly insecure.[63] Mr Holmes did not select a random member of the public but instead framed a criminal who had committed offences against children. The public may conclude from this case that to the extent that the police frame people, they choose people who have committed similar offences in the past.

It would seem that the question of whether Mr Holmes should fabricate evidence is dependent upon an assessment of the damage which would be caused if the fabrication is exposed, the probability of such exposure occurring and the benefits of fabrication if there is no exposure. However, Pettit would argue that while his approach gains in strength if there are no cases in which republican liberty could be promoted by violating rights, his approach still stands even if such cases exist. This is because the decision on whether to respect rights should not be made on a case-by-case basis, but rather, a decision should be made in advance to treat certain rights as inviolable. This will ultimately promote republican liberty to a greater extent. Pettit might say to Mr Holmes:

You claim that manufacturing evidence here promotes liberty. Even if your calculations are correct in this case, you should still not manufacture evidence against Mr Smith. If you calculate here, you will be calculating in other cases. While the risk of being caught on any particular case may be small, the risk of being caught in relation to a case, would be larger. Furthermore, because cases such as Mr Smith's are exceptional, little liberty is gained by adopting calculation as a general approach, while much liberty is placed at risk.

However, Mr Holmes might reply that he recognises certain rights as presumptively valid and would only engage in an extended process of calculation in exceptional cases such as the present one. To be sufficiently alert to recognise an exceptional case does not require a calculative approach to each case. Mr Holmes may say that he has not fabricated evidence previously and would not expect to face another situation in which fabrication is warranted. This suggests that Mr Holmes could argue persuasively that the fabrication of evidence in this case furthers the liberty of the community, excluding Mr Smith.

Turning to Mr Smith, his imprisonment would involve procedural arbitrariness. Procedural arbitrariness occurs where a decision-maker enjoys an unchecked discretion to interfere with another member of the community.[64] Since Mr Holmes' decision to manufacture evidence must be secret if it is to succeed, it involves an unchecked discretion. Therefore, Mr Holmes is dealing with a trade-off between Mr Smith's intensity of liberty and the intensity of liberty of the community, in particular, children and their intimate associates. One way of resolving this in favour of Mr Smith is by privileging objective over subjective liberty. Mr Smith suffers an objective diminution in liberty which does not translate into the subjective suffering one might expect. On the other hand, the community suffers subjective insecurity which is much greater than their actual vulnerability. Pettit assumes that psychological states will generally reflect actual vulnerability to interference,[65] so this might suggest that we can generally pursue policies which focus on objective liberty. However, this does not seem a defensible approach here since we know that the assumption does not hold and, furthermore, a distinguishing feature of Pettit's republican liberty is that it is deeply concerned with psychological liberty.

How then, do we resolve this trade-off? I mentioned, in Part II A, Pettit's view that intensity of freedom is best optimised through its equal distribution but those who are unconvinced of this may cast the goal of society as equally intense liberty. Pettit does not discuss in detail how trade-offs are to be made between groups in order to achieve this goal. However, he has suggested that the Gini index of economic inequality, which focuses on the statistically expected difference in economic fortunes within a randomly chosen pair of people in a society, might be useful.[66] This is sensitive to the numbers at less than the best level as well as to differences between levels. This sensitivity to numbers leaves it open for Mr Smith's liberty to be sacrificed for the sake of the liberty of children and their intimate associates. For a firmer basis for protecting Mr Smith, we must look beyond republican liberty.

B. The Need for Desert

An important value which is lacking from Pettit's approach is desert: a powerful reason why it is wrong to punish the innocent is because such punishment is undeserved. We would say here that a fair investigative process is necessary for a fair trial, which is itself necessary to protect the right of the innocent not to be convicted. This right is derived from the more general principle that punishment should only be imposed if it is deserved and whether it is deserved depends upon culpability.

However, Braithwaite and Pettit reject this principle and argue instead that punishment should only be imposed when this promotes liberty. Punishment might achieve this through incapacitation and rehabilitation (both reduce the chance that the offender will re-offend)[67] and through deterrence and moral education[68] (both reduce the chance that the offender and others will commit crimes). Their main justification for punishment is reprobation, a form of moral education which ensures that criminal behaviour is subject to disapproval.[69] In the case of Mr Smith, a substantial prison term would indicate the heinousness of the offences. Clearly, Mr Smith will not only receive this message but also the negative message that even the police are prepared to violate the law. However, as far as the community is concerned, the effectiveness of reprobation is not dependent upon actual but only upon perceived guilt. Reprobation may be best achieved by punishing Mr Smith since if no one is convicted, the criminal justice system fails to express disapproval of some appalling crimes and perhaps conveys the unfortunate message that one can get away with even the worst crimes.

In criticising crime preventionism, Braithwaite and Pettit say:

If there is reason to incapacitate or rehabilitate an actual offender, there is equal reason to provide such treatment for anyone who is judged likely to offend. And if there is deterrent reason to penalize an actual offender, so there may be reason to penalize any party - in particular, any potential offender - who is widely believed to be guilty of a crime, or who can be made into a credible scapegoat.[70]

They regard their own use of the preventionist aims of incapacitation, deterrence and rehabilitation as immune from these criticisms because they are not merely concerned with preventing crime but are deeply concerned with people's sense of security. However, my example of the detective suggests that a concern with people's sense of security is consistent with violating a right which protects the innocent. Even with security as one's target, the goals of reprobation, deterrence and incapacitation may suggest imprisoning Mr Smith.

Braithwaite and Pettit's approach, like crime preventionism, relies upon empirical contingencies in protecting rights. In most cases, a calculus based on republican liberty will suggest that the rights of the innocent should be respected. However, the differences that flow from a distinction between innocence and guilt as they are caught by the calculus seem insufficiently substantial. A decision on whether to lock up Mr Smith hinges upon an empirical assessment of how likely it is that his innocence would be publicly revealed and what the public's reaction to this would be. It excludes the powerful intuition that the conviction of the innocent is manifestly unjust because it is undeserved. Furthermore, there are other values which could dissuade Mr Holmes from manufacturing evidence. Public officials are entrusted with substantial powers by the community, and accountability for the exercise of those powers requires their honest exercise. In agreeing to become a public official, that official gives a commitment to follow legal procedures.

Issues of trust, accountability, honesty and faithfulness to commitments undertaken are involved. It is plausible to suggest that while Mr Holmes was unconvinced by republican liberty, he may be convinced by other values that he should not manufacture evidence.

As mentioned before, Braithwaite and Pettit reject desert as a justification for punishment partly because they are concerned that this plays into the hands of conservative law-and-order politicians who favour oppressive punishments.[71] However, while desert is open-ended in the sense that some people use this value to justify punishments that we may find oppressive, consensual judgments about the relative seriousness of offences can limit punishment.[72] This is because of the community's belief that punishment should be roughly proportional to the seriousness or culpability of the offence in question. The most plausible explanation for this belief is that punishment should be in accordance with desert. I have also mentioned that desert-based theorists argued in the 1970s that punishment should be proportional to culpability and used this notion to oppose the oppressive punishments which were being imposed under the reigning theory of preventionism.[73] One difficulty with preventionism can be illustrated with a hypothetical. Suppose that burglars can be deterred more effectively through severe punishments than murderers, and burglars are more likely to re-offend than murderers. Preventionism might suggest penalising burglary more severely than murder. However, desert allows us to resist this through the claim that it is generally unfair to punish burglars more than murderers since murder generally involves greater culpability.

Braithwaite and Pettit would oppose harsher penalties for burglars by arguing that such penalties could undermine confidence in the criminal justice system: such penalties suggest that parliament does not regard taking the life of another as the ultimate evil.[74] They are relying on reprobation, that is, the role of penalties in communicating to the offender and the community the seriousness of particular offences. However, punishment is indicative of seriousness because people generally believe that penalties should be proportional to seriousness. I have just suggested that this belief is based on desert. In this way, Braithwaite and Pettit themselves rely upon the community's sense of desert to show how a republican system would avoid penalising burglars more severely than murderers. On the other hand, if the community does not believe in desert, Braithwaite and Pettit cannot oppose harsher penalties for burglars.[75]

What about people who believe on the basis of desert that penalties in general should be increased to a level we may regard as excessive? We cannot rely on a consensus on the ranking of offences to resist this. Braithwaite and Pettit would seek to persuade these people to abandon their commitment to desert and recognise that liberty is best promoted with lesser penalties. However, this may simply alienate them since their commitment to desert may be strongly held. A more persuasive approach would be to accept their commitment to desert, but then argue that lesser punishments overall are consistent with desert. There is no simple correspondence between a particular offence and a particular level of punishment.[76] Should a maximum penalty be, say, 20 years or 30 years imprisonment? Given that desert leaves this indeterminate, the lower figure could be recommended since it promotes other goals, perhaps including republican liberty. Furthermore, a consideration of situational and other factors might suggest that offenders are generally less culpable than one might have assumed.[77]

Braithwaite and Pettit's opposition to desert lies not just at the political[78] but also at the theoretical level. They cogently criticise some rationales which have been proposed for the principle of desert in punishment.[79] However, they do not provide a convincing argument against intrinsic negative retributivism. Intrinsic negative retributivism suggests that it is intrinsically wrong to punish the innocent or impose a punishment which is disproportionate to the criminal's culpability.[80] It is true that the intrinsic nature of the principle may cause some unease. While all moral reasons eventually rest upon fundamental intuitions, it may be disconcerting to reach such intuitions almost immediately since this halts the process of reasoning backwards. This can be illustrated with the following exchange:

Qn: Why is it wrong to punish Mr Smith?
Ans: Because it is wrong to punish the innocent. Qn: Why is this wrong?
Ans: Because it is undeserved.
Qn: Why is it undeserved and why is this relevant?
Ans: It simply is undeserved. As a matter of basic intuition, I believe that punishing the innocent is a gross injustice. Punishment should only be imposed on those who are culpable and should not be excessive, given their level of culpability.

Nevertheless, if we are committed to this intuition, can find no convincing derivation of the intuition in terms of other values and there are no contrary intuitions which can shake our confidence in it, then it is reasonable to give the response quoted above. It represents a bedrock intuition which is, surely, widely shared in our community.[81]

Those who favour consequentialism may be concerned, though, about accepting intrinsic negative retributivism since it is almost always understood as a deontological principle. Indeed, Braithwaite and Pettit say:

Retributivists may argue against us that in deriving the right of the innocent not to be punished, and the limit on punishment of the guilty, from our republican goal, we make those rights and those limits too contingent; they cease to be properly moral constraints. We would reply that if a right or a time limit is to be honoured generally, it must be made ffnerally intelligible and that the best way to do this is a consequentialist derivation.[82]

However, they do not seriously consider the possibility of intrinsic negative retibutivism being a consequentialist target.[83] I will give a brief indication of how desert could be included within a consequentialist framework and an advantage of doing so.

C. Placing Desert and Other Values within a Consequentialist Theory

When negative retributivism is understood deontologically, agents are simply not permitted to impose a penalty disproportionate to culpability. However, this constraint can be problematic. Suppose that imprisoning an innocent person is the only way to prevent many innocent people being killed.[84] Such a loss of life would also offend the principle of desert: while negative retributivism is directed to the question of punishment, it is based on desert and therefore more generally reflects opposition to undeserved hardship or loss.[85] Furthermore, the deaths of innocent people would appear to be a greater violation of desert than that involved in imprisoning an innocent individual. In deciding whether to imprison or not, it seems odd to exclude from consideration consequences which involve a greater violation of desert.[86] This difficulty suggests that we should seriously consider the possibility of including desert within a consequentialist framework.[87]

Desert is not a comprehensive value so it must be placed within a pluralistic consequentialist theory. The most elegant pluralistic theory involves lexical priority: one must satisfy one value, before considering the next.[88] We could firstly promote desert, before proceeding to other values such as republican liberty and utility. However, even if one prioritises desert, it is not clear that Mr Smith's interests must be vindicated. It might be argued that the insecurity suffered by not convicting Mr Smith is also undeserved and this injustice, considering the number of sufferers, outweighs the injustice Mr Smith would suffer. This conclusion can be avoided with Rawls' principle of giving priority to the worst-off.[89] However, I agree with Scheffler that this is rather rigid and a distributive principle which gives greater weight to the worst-off is preferable.[90] This begs the question of how much greater weight, but I think that this is best resolved contextually. We would say that the violation of desert involved in imprisoning an innocent person is very much greater than that involved in being subject to the insecurity of believing a murderer is on the loose so the former should be given a weighting to ensure that the individual's interest in not being unfairly imprisoned prevails.

It might be argued that this method of treating desert involves an approach which is not significantly different from Pettit's: in the consequentialist calculus, I have merely replaced diminution in liberty with undeserved hardship. To the extent that desert increases our concern for Mr Smith, it also increases our concern for the community since the community will suffer not just insecurity but undeserved insecurity. What have children done to deserve living in fear? The difficulty with this argument is that it assumes that desert increases the interests of Mr Smith and the interests of the community by the same proportion. However, the distributive principle I suggested is a flexible one: it allows for different weighting to be assigned to the worst-off as different values are considered. Imprisoning an innocent person for grave offences which they did not commit strikes at least most of us as an injustice which cannot be outweighed by some undeserved insecurity that the community would otherwise suffer. If we exclude desert and only consider the issue in terms of republican liberty, we may feel more comfortable sacrificing Mr Smith and may therefore give his interests a lower weighting.

It can also be argued that desert as a consequentialist target provides weaker support for the right not to be framed than we may have supposed, which suggests that we in fact treat desert as a deontological constraint. It is true that desert only offers absolute support for the innocent not being punished if we focus upon the individual victim and do not consider what violations to the principle of desert are wrought upon others by failing to punish that individual.[91]

However, this exclusion of considerations seems arbitrary, so rather than adopting a deontological approach, it seems better to consider whether there are other values which help to explain why we object to Mr Smith being framed. I mentioned in Part III B that the values of trust, accountability, honesty and faithfulness to commitments undertaken may be compromised by framing Mr Smith. Having desert as a consequentialist target has the advantage of forcing us to consider deeply what other values are implicated by convicting the innocent and whether we can explain our strong intuitive objection to this in a rational manner, without rigidly excluding considerations through the use of agent-centred constraints.

The fact that there are other values which favour Mr Smith tells against giving lexical priority to desert. Such priority leaves open the odd result that while desert plus other values may be sufficient to vindicate Mr Smith's interests when taken together, his interests should nevertheless be defeated because neither desert nor other values are sufficient by themselves to vindicate his interests. Admittedly, lexical priority also excludes values such as republican liberty which might suggest framing Mr Smith. However, I think the values favouring Mr Smith are sufficient to withstand the contrary claims of other values. The problem with Pettit's approach is that it excludes values which can tilt the scales in favour of Mr Smith.

Finally, it can be argued that while placing desert within a consequentialist theory avoids the deontological paradox, it does not avoid complexity. My approach involves multiple values and suggests that the weighting of values must be determined contextually in the absence of a precise and general rule. Furthermore, negative retributivism is not a comprehensive theory of punishment, let alone of the criminal justice system. Desert also suggests punishment of those who are culpable at a level commensurate with their culpability[92] and I have not considered how this intuition is to be accommodated with, say, considerations based on republican liberty.[93] The approach I favour is complex, but as Braithwaite and Pettit point out, it is not complexity but excessive complexity which is problematic.[94] If the inclusion of desert and other values is essential (as I have argued), the complexity introduced by their inclusion cannot be regarded as excessive. I return to the issue of complexity in Part IV C.

D. The Relevance of the Case

I will now deal with some objections which may be raised to my police detective hypothetical. It might be claimed that its highly contrived nature suggests that in most cases republican liberty will provide sufficient support for uncontroversial rights. Furthermore, we can create a hypothetical in which desert fails to support such rights. Suppose that before Mr Smith was convicted for his previous offences, he had admitted to Mr Holmes that he had also committed a much more serious offence. However, there was insufficient evidence to secure a conviction on this. Assume also that framing Mr Smith for the offences committed by Mr Jones would lead to a punishment which is roughly equivalent to the punishment Mr Smith avoided due to one of his offences not being proved. We may decide that desert is served here by framing Mr Smith.

I concede this, but my argument is not that it is statistically more likely that desert will support the right not to be framed in the various unusual circumstances that we can imagine. The reason for considering the right not to be framed is because in general terms it is a right which we are deeply committed to. It represents one of the fixed points against which we judge moral theories.[95] However, while we are deeply committed in general terms to this right, there may be unusual circumstances where that commitment weakens. I would suggest, though, that while my original hypothetical is unusual, it still involves a situation where our commitment remains strong. I demonstrated that this commitment cannot be understood purely in terms of republican liberty. The fact that the hypothetical is unlikely to occur is beside the point. The hypothetical is merely a means of demonstrating the fact that Pettit's theory does not reflect the depth of, and the reasons for, our commitment to the right and this is relevant to more usual circumstances. We judge a moral theory not just on the basis of whether it generally supports attractive outcomes, but on whether, in reaching these outcomes, it does so for the right reasons.

Turning to the modified hypothetical in which Mr Smith has escaped punishment for an earlier offence, the fact that desert may not support the right here does not mean that there is something wrong with desert and therefore desert should be discarded. Instead, as proposed in Part III B, we should search for other relevant values. I mentioned then some values, such as trust, accountability, honesty and faithfulness to commitments, which may support adherence to legal procedures. Those values, perhaps together with other values, might outweigh contrary intuitions based on desert, and lead to the conclusion that Mr Smith should not be framed. On the other hand, some may feel that other values do not outweigh desert here and that Mr Smith should be framed. In that case, they must accept that their support for the uncontroversial right not to be framed is not absolute. My main point is not that uncontroversial rights must always be upheld, but instead, that there are important values which should not be replaced by republican liberty. The fact that my approach leaves open the possibility that manufacturing evidence can be justified may cause some disquiet. Manufacturing evidence does not, after all, only occur in philosophers' hypotheticals, but in real life. However, in so far as we engage in moral discussion with police officers, my approach would, I think, provide a more persuasive case against manufacturing evidence. I also mentioned in Part III B that to suggest to people that they exclude desert from consideration may alienate them, for desert is often a deeply held value. It is surely better to acknowledge that desert can sometimes appear to favour the violation of legal procedures (republican liberty does not differ in this respect). While it may be more comfortable to regard such violations as simply evil actions which must be motivated by base emotions or interests, the truth may be more complex. We may have to admit that such actions can be motivated by genuine conceptions of what lies in the public interest. It is .then our task to bring the full panoply of values and arguments that we can to explain why our inclinations, including moral inclinations, to violate procedures should be resisted. A proper understanding of the public interest suggests extremely strong grounds for upholding legal procedures. Furthermore, it might be helpful to increase police awareness of how their involvement in investigations can colour their judgment and impose pressures upon them.

Another objection is that my hypothetical does not relate to judicial decision-making. One might contend that even if republican liberty does not provide the police with sufficient reason to respect rights, it may be adequate for judges. While Mr Holmes may secretly falsify evidence, judicial proceedings are open and recorded and judges must provide reasons for their decisions. This precludes judges from saying one thing publicly while privately doing something else.[96] However, compelling judges to give reasons does not compel them to reveal their actual reasons. Assume that Mr Smith elected to have a judge rather than a jury determine whether he is guilty. The concept of reasonable doubt has some vagueness. If it is interpreted very narrowly, few if any convictions would be obtained in contested matters, but if it is interpreted too broadly, many innocent people may be convicted. Tribunals of fact enjoy some discretion over what is reasonable doubt. If a judge applies republican liberty to the interpretation of reasonable doubt, the factors which convinced Mr Holmes to manufacture evidence would also tell in favour of a broad interpretation of reasonable doubt. Again, in this situation, a consideration of desert may weigh heavily in favour of the accused.

IV. THE CASE OF DIETRICH v R

A. The Case's Impact upon Republican Liberty

I will now focus directly on judicial decision-making, using the case of Dietrich[97] as an example. In this case, the High Court held that the right to a fair trial generally requires the provision of legal representation to accused persons who are charged with serious offences and cannot afford representation.[98] A court should stay proceedings until the accused is provided with representation. The previous position was that judges should adjourn proceedings to enable an accused to apply for representation, but once the accused had exhausted the avenues for obtaining representation, the trial should proceed even though the accused was unrepresented. The position now is that there will generally be no trial unless the accused is provided with representation.

I mentioned in my introduction that Braithwaite, in his model of judicial deliberation, suggests that at least in hard cases concerning criminal justice, judges could consider which outcome best promotes republican liberty.[99] I will now consider whether republican liberty provides strong support for the judiciary finding in favour of Dietrich. As with the detective case, we are dealing ultimately with the right not to be convicted of offences one has not committed, or more specifically, with the right to a fair trial. One of Braithwaite and Pettit's criteria for an acceptable ethic theory for the criminal justice system is that it sufficiently motivates respect for uncontroversial rights.[100] This includes the right to a fair trial. Nevertheless, some elements of a fair trial are more controversial than others. Being subject to an investigative process which is aimed at discovering rather than manufacturing evidence is relatively uncontroversial. However, the question in Dietrich was whether the Court should declare that a fair trial involving serious criminal charges required legal representation. While there may be little doubt that legal representation leads to fairer trials, this does not mean that its absence violates a minimum acceptable standard of fairness; whether it does so is a controversial question. If republican liberty fails to sufficiently support the 'right' to the provision of legal representation which the High Court found,[101] this does not, then, constitute the serious failing that occurs when republican liberty fails to adequately justify an uncontroversial right. On the other hand, Braithwaite and Pettit would surely support the Dietrich decision: it reflects the degree of concern for accused and guilty persons that they encourage and, furthermore, they are not averse to some judicial activism.[102]

There is another difference between Dietrich and the detective case. The detective case tested Pettit's claim that republican liberty avoids the utilitarian problem of approving the violation of important, uncontroversial rights when this seems intuitively wrong. Dietrich provides a way of testing Pettit's claim that republican liberty provides a sufficient basis for deciding the content of rules. I begin by considering Dietrich in terms of republican liberty.

Dealing firstly with the positive consequences of finding in favour of Dietrich, he was only offered legal assistance if he pleaded guilty, thereby depriving him of the opportunity to argue his innocence with the benefit of legal assistance, which was his desire. With legal representation, his prospects of a jury finding additional counts in the charge not proved and consequently obtaining an acquittal or a reduced sentence would probably have improved.[103] Having legal representation would also have given Dietrich some sense of security both before and during the trial. Undoubtedly, Dietrich's prospects of republican liberty, and the prospects of others finding themselves in a similar position, were enhanced by the decision. Turning to the decision's impact upon the general community, many people would not have heard of the decision and are unlikely to become aware of it. Of those who are aware of the decision, most would not anticipate ever being in Dietrich's position. The decision may only have improved the sense of security of a small number of people.

While the judges could have decided that the right would improve the sense of security of an indeterminate but possibly small number of people, calculating the costs of the right would have been more complicated, since the judges would not have known how the Commonwealth Government and State Governments would react to their judgment. This is especially so since the Commonwealth argued in the case against the Court finding such a right. If the Commonwealth or States refused legal assistance so that accused persons avoided trials, this would generate considerable publicity and could make many people insecure. This very insecurity, though, might suggest that no government would allow this to happen. However, given the controversial nature of the demarcation of responsibility over legal aid between the Commonwealth and States, one cannot be sure of this: a stand-off could occur.[104] Nevertheless, the more likely response is not to refuse legal aid where a stay is ordered, but instead to direct that accused persons facing serious charges be accorded greater priority in legal aid funding; this is in fact what Mason CJ and McHugh J envisaged.[105] However, while many people may not expect to ever be charged with a serious offence, they may regard involvement in civil proceedings such as family law matters as a possibility. They may be concerned about how well they can protect their own and their children's interests if legal aid funding is shifted more towards serious criminal cases. In deciding what best promotes aggregate liberty, judges would face a complex task plagued by uncertainty. Mason CJ and McHugh J admitted that they did not have the information to assess the full practical implications which would flow from adjourning trials.[106]

Braithwaite and Pettit admit that uncertainty plays a large role in the application of consequentialist theories, unlike deontological theories, since we often cannot predict consequences. However, they suggest that this uncertainty will work against government intrusion because while every intrusion such as the criminalisation of conduct, prosecution or punishment does immediate and unquestionable damage to someone's liberty, the benefits provided by these initiatives are almost always of a probabilistic character.[107] They say it is therefore clear that the onus of proof should lie on those seeking to justify the intrusion. Applying this to Dietrich, we are concerned with a right that prevents the intrusion constituted by subjecting individuals who cannot obtain the protection provided by legal representation to a criminal trial involving serious charges. This right definitely improved the prospects of liberty of Dietrich and those in a similar position to him. I have suggested that the adverse impact upon liberty is uncertain. Braithwaite and Pettit would therefore support the right.

B. Judicial Activism

Pettit's approach does provide a moral justification for a right to legal representation. However, those who support the Court's decision should be concerned about whether republican liberty provides a sufficiently strong justification for the decision. Especially in the context of judicial reasoning, it is important that a comprehensive moral theory not only supports the right but also does not exclude arguments which can strengthen the justification for finding the right. Judges are confronted with the view that controversial changes to the law, especially those with fiscal implications for the government, should be left to parliament. Courts may feel that it is not worth risking their authority on legal changes which are only on balance beneficial as opposed to being imperative. More generally, judges may be deferential when the case in favour of a right is based on considerations which parliament may be better able to determine. If a moral theory relies on these factors and excludes considerations which judges are well placed to determine, judges applying this theory may be reluctant to find the particular right at issue.

Theorists who favour the activism involved in finding rights tend to draw attention to deficiencies that the parliamentary process is more prone to than the judicial process. Michelman, for instance, regards the behaviour of the government and parliament as characterised to a significant extent by pluralist behaviour. With pluralism, participants negotiate with others in order to reach a deal which best promotes their own interests, narrowly conceived.[108] This contrasts with the republican ideal of public-spirited citizens deliberating and acting together; I mentioned in Part II A that unlike Pettit, Michelman associates republican theory with positive liberty. Michelman suggests that the isolation of the judiciary from the pluralist pressures that politicians face places judges in a good position to engage in republican deliberation.[109] This in turn suggests that on issues in which the public interest diverges most sharply from pluralist outcomes, judges have a special responsibility to produce outcomes which reflect republican deliberation. The rights of accused persons charged with serious offences may be an example: accused persons are relatively few in number and attract limited public sympathy.

However, judges are unlikely to respond to such cases in an activist manner unless the moral theory they apply contrasts with pluralism. It might be argued that in adopting an aggregative approach, Pettit's theory is not sufficiently distinctive from pluralism, since pluralism also involves the aggregation of people's interests, albeit an aggregation reflecting an uneven spread of political resources. It is in order to sharply distinguish political from judicial decision-making that Dworkin limits judges to deontological reasoning.[110] However, while there must be a significant difference between pluralism and judicial reasoning, this difference need not be extreme.[111] While Pettit's approach involves aggregation, it is clearly distinguishable from pluralism.[112] The difficulty with Pettit's approach is that it fails to provide judges with the strongest possible case in favour of the right. I mentioned that Pettit relies on uncertainty to suggest that his approach works in favour of results which protect individuals against government intrusion. Such uncertainty could, however, undermine judges' willingness to engage in the activism which finding new rights involves. Judges might think that the uncertainty in the costs of a particular right is largely due to the limitations of courts. A relevant limitation here is ignorance of how governments would react to courts ordering stays of proceedings. With republican liberty emphasising the importance of people's feelings, difficult empirical questions are encountered, questions which parliament may be better able to resolve.

A stronger case for activism could be made out if, rather than relying on uncertainty, judges could say that even if the worst plausible consequences were to occur, a right to representation should still be found. It will come as no surprise to the reader that I believe that desert is helpful here. The Dietrich right has an important part to play in minimising the chance of injustice, the injustice of an undeserved punishment. Of course, judicial recognition of the fact that justice is at stake here does not necessarily motivate a commitment to finding the right. Dawson J, who dissented along with Brennan J, stated that justice cannot be pursued in isolation: there are competing demands upon the public purse and it is for parliament to balance these demands.[113] However, the injustice of an undeserved punishment is one which perhaps most people are concerned to avoid. It might be the case that desert is what convinces judges to support the Dietrich right. It should be considered along with other relevant values.

It is interesting to note that while the majority judgments affirm the importance of justice for accused persons, attention is still given to other considerations. Justice Gaudron's statement that: "whatever the consequences and whatever the cost [of finding this right], it is for the courts to decide what is or is not fair in a criminal trial",[114] affirms the primacy of the value of justice in judicial deliberation. While the other majority judges are not as explicit as Gaudron J on this, I think that this approach is implicit in their judgments. The judges do not, however, entirely ignore other considerations. While the above quotation from Gaudron J might suggest that fairness operates as a deontological constraint, she does say that finding a right to legal representation may not be appropriate in third world countries due to their limited financial resources.[115]

She is implying that for developed countries such as Australia, the right is not too costly. Mason CJ and McHugh J proceeded on the footing that the right would not impose a substantial burden upon the government.[116]

C. Republican Deliberation

The approach I favour can be elaborated upon by referring again to Michelman's model of judicial deliberation. As mentioned before, this is based on the ideal of virtuous citizens seeking together to ascertain the public good.[117] Michelman does not define the public good in substantive terms such as maximising liberty, but in procedural terms: the public good is whatever would be- agreed to by all interested individuals as defining the public good. He sees republican deliberation as involving elements in tension which we reconcile through the exercise of judgment.[118] We need to reconcile, for instance, our need for rules that limit discretion and our need for judgments which appropriately reflect contextual differences. In defining the public good in terms of what is consensually decided, Michelman is not suggesting that consensus is generally achievable. Instead, consensus is a regulatory ideal; it emphasises the importance of sensitivity and openness to the rich plurality of perspectives in the community. It suggests a preparedness to engage in critical reflection on existing rules and practices; only through this can we hope to discover the public good. While Michelman relies on a diversity of perspectives to provoke critical reflection, he recognises that republican deliberation is dependent upon the participants enjoying some commonality. Without this, everything would be open to discussion with little chance for agreement. Republican deliberation requires diversity and commonality, elements which exist in tension.

While he suggests that republicanism can provide substantive guidance and accepts that substantive theories can illuminate moral issues, he does not recommend any comprehensive substantive theory.[119] By contrast, Braithwaite and Pettit propose a single criterion to determine all issues concerning criminal justice.[120] Pettit goes further in his recent book, suggesting republican liberty as a value to guide all policy-making. From Michelman's perspective, one would be concerned that Braithwaite and Pettit exclude considerations which may be important in dealing with particular situations.

However, stripped of its claim to comprehensiveness, Pettit's theory can be welcomed as one which illuminates moral issues. With my detective hypothetical I suggested that arguments based on republican liberty may be cogent. They may also offer grounds for opposing oppressive punishments. With respect to Dietrich, desert fails to account for all the reasons why the Court's decision should be supported. The decision not only reduces the likelihood of unjust convictions but also establishes a desirable relationship between the state and citizens. Pettit's vision of citizens who feel secure and can look others in the eye is an attractive one, and it applies not only to citizens' relationships with each other and with private organisations but also to their relationships with the state. This ideal is subverted in a criminal trial, especially one involving serious charges, where the accused is unable to obtain representation and lacks legal expertise. Here the state, backed by its overwhelming coercive powers and substantial resources, articulates a case through legal representatives, using rules which the accused has little understanding of. It is not difficult to imagine the fear such a process can instill, the sense of subordination involved in being at the mercy of whatever crumbs of assistance the prosecution and the judge provide.[121]

This image of vulnerability and subordination may be sufficiently disturbing to commit some people to the public provision of legal assistance. For others, the image which motivates this commitment will not be the indigent accused, trembling in the dock, trying desperately to defend himself or herself through a bewildering process. Instead, it may be that of the innocent individual, languishing in jail, angry and depressed at the injustice that has been perpetrated. This second image does not have an equally powerful equivalent in discourse based on republican liberty; it derives its strength from the value of desert. Michelman's model encourages us to consider all these images. Together, they might galvanise commitment to the right to representation even if such a right has substantial costs.[122] We should not only consider the consequences of finding the right but also consider arguments suggesting that not all consequences may be relevant to the judiciary. Even if the judicial creation of a right does lead to government policies which are negative, this does not mean that judges must feel responsible for all these consequences. Governments must take responsibility for their actions: they can choose not to regard judicial endorsement of a right to representation as a nuisance to be accommodated within existing funding, but rather as a positive development which can legitimise additional funding of legal representation or which can prompt a rethink into dispute resolution procedures.

While I have suggested that Pettit's approach excludes important considerations, it might be argued that it at least has the virtue of simplicity; it provides a clear consequentialist target based on one value, avoiding the complexities involved in pluralistic targets. Michelman' s model not only stresses the importance of diversity or openness but also commonality or closure. The issues on which we disagree must be sufficiently narrow to be manageable; perhaps monistic consequentialism based on republican liberty provides us with the discipline we need. However, in deciding what form of punishment is appropriate for a particular crime, Braithwaite and Pettit suggest a wide range of considerations including the preventionist goals of deterrence, incapacitation and rehabilitation, some communitarian approaches to punishment,[123] and even the community's sense of desert. While republican liberty is suggested as a simple consequentialist target, Sadurski argues convincingly that Braithwaite and Pettit's target accommodates a mixed bag of distinct concerns, leaving us with difficult and painful choices in reconciling divergent considerations.[124] Braithwaite and Pettit's approach to criminal justice is not so simple yet it still excludes important values.

It might also be argued in Pettit's defence that while he presented republican liberty as the sole criterion for a criminal justice system, in his recent work, Republicanism, he says:

There is nothing in this book to support explicitly the traditional assumption that ... [republican liberty] is the only goal with which our political institutions need to be concerned. But my own view is that once we fully understand the demands of promoting [republican liberty] ... we shall find this claim quite congenial.[125]

He also says that the ambition to provide the political philosophy to end all political philosophies is unrealistic since the conversation of politics is constantly evolving and shifting as new languages come to the fore.[126]

Nevertheless, in the postscript to the forthcoming edition, Petit stands by his suggestion that the common good can be determined by reference to republican liberty alone.[127] Indeed, a major justification for his work on republican liberty relies upon this being the sole value of a political system. He says in his book that while his theory points in a direction similar to left-of-centre liberals, it starts from a base that is less contentious. While left-of-centre liberals have to rely on controversial values such as equality which right-wing liberals may reject, Pettit offers those with left-of-centre inclinations a common ground for arguing with rightist opponents who give priority to safeguarding liberty.[128]

Admittedly, it is a particular conception of liberty, a conception which differs from that associated with liberalism. However, Pettit says that the republican conception should be attractive to people of various political persuasions.[129] It provides a common language for political discourse which accommodates the concerns of diverse ideologies including liberalism, socialism, feminism and environmentalism, while at the same time disciplining the types of arguments which can be presented. While Pettit does not put it this way, he seems to be proposing his theory as one which provides hope of greater agreement in favour of a left-of-centre political agenda. In contrast with current debate with its messy multiplicity of controversial values, we can have a single, relatively uncontroversial value as our lodestar. Pettit is not, of course, placing before us the miraculous vision of a community harmoniously following the same path. He acknowledges the limited impact philosophy generally has on political debate. Furthermore, republican liberty provides ambiguous guidance, leaving much room for debate and disagreement. There can be little doubt, though, that Pettit is presenting republican liberty as the sole goal for our political system and justifies his theory on this basis.[130] His recent book suggests no retreat from the claim made in Not Just Deserts that republican liberty should be the sole value of the criminal justice system. Instead, the claim has been extended to cover political issues in general.

V. CONCLUSION

Braithwaite and Pettit have used the latter's theory of republican liberty in a sophisticated attempt to provide a comprehensive account of how we should reason on issues concerning the criminal justice system. Pettit's theory not only involves an interesting interpretation of republican liberty but makes this value the basis of a consequentialist moral theory, thereby engaging with the fundamental divide within ethics, between deontology and consequentialism. By making republican liberty the sole goal of the criminal justice system, it challenges both crime preventionism and retributivism. I have considered how effectively his theory promotes adherence to an uncontroversial right and supports a more controversial right that the High Court has found. I argued that his approach excludes important values and that a pluralistic consequentialist theory would be more attractive. Braithwaite and Pettit's approach fails to avoid some of the difficulties of crime preventionism and it is unconvincing in arguing that we can dispense with retributivism.

While I focused upon some important issues concerning criminal justice, my discussion has relevance to Pettit's recent book. First, I relied upon its exposition of republican liberty, and secondly, my discussion of criminal justice may point to difficulties which also plague the application of republican liberty to the wide range of issues covered in this book. This paper raises the question of whether, on some of these other issues, Pettit may also have failed to strike a satisfactory compromise between commonality and diversity.[131] In constructing an entire theory around republican liberty, and considering how it applies to policy issues and broad political questions, Pettit has undertaken an ambitious project which certainly merits attention. Whether he has succeeded in establishing republican liberty as the supreme value upon which political and legal discourse should be based is more doubtful. At least in the area of criminal justice, this project seems to have foundered.


[*] BA LLM (Hons) (Syd). PhD student, Law Faculty, University of Sydney. I would like to thank Margaret Allays, John Braithwaite, Roger Magnusson, Philip Pettit, Fenella Thomas, Wojciech Sadurski and this journal's two anonymous reviewers for their helpful comments.

[1] F Michelman, "The Supreme Court, 1985 Term - Foreword: Traces of Self-Government" (1986) 100 Harv LR 4; and "Law's Republic" (1988) 97 Yale LJ 1493.

[2] C Sunstein, "Beyond the Republican Revival" (1988) 97 Yale U 1539; and The Partial Constitution, Harvard Press (1993).

[3] There is some consideration of republican theory in the context of Australia becoming a republic in (1993) 28 Australian Journal of Political Science (Special Issue: Australia's Republican Question); "Debating the Question of Republicanism for Australia: A Symposium" (1992) 6(2) Legislative Studies 21; W Hudson and D Carter (eds), The Republicanism Debate, UNSW Press (1993); B Galligan, A Federal Republic, Cambridge University Press (1995) pp 16-17 and generally.

[4] G Williams, "A Republican Tradition for Australia?" (1995) 23 Fed LR 133 at 143 argues that it is not part of our constitutional tradition. The following writers disagree: A Fraser, "In Defence of Republicanism: A Reply to George Williams" (1995) 23 Fed LR 362; J Williams, "`With eyes open': Andrew Inglis Clark and our republican tradition" (1995) 23 Fed LR 149 and "Revitalising the Republic: Lionel Murphy and the Protection of Individual Rights" (1997) 8 Public LR 27; and J Warden, "The Fettered Republic: The Anglo-American Commonwealth and the Traditions of Australian Political Thought" (1993) 28 Australian Journal of Political Science (Special Issue) 83. The controversy is partly dependent upon which interpretation of republicanism is adopted. G Williams says at 148 that Pettit's interpretation is closer to our tradition than Sunstein's, although a significant gap remains.

[5] P Pettit, "Liberalism and Republicanism" (1993) 28 Australian Journal of Political Science (Special Issue) 162.

[6] J Braithwaite and P Pettit, Not Just Deserts: A Republican Theory of Criminal Justice, Oxford University Press (1990). I am not implying that Pettit, having formulated his understanding of republican liberty, decided to apply this to the criminal justice system. Instead, Braithwaite and Pettit together decided that resilient non-interference was an appropriate consequentialist target for the criminal justice system, and later David Neale suggested to them that this understanding of liberty is republican. Pettit then pursued this suggestion.

[7] J Braithwaite, "Community Values and Australian Jurisprudence" [1995] SydLawRw 21; (1995) 17 Syd LR 351.

[8] Ibid at 368, 371. There have been several critiques of Braithwaite's model: K Ziegert, "Judicial Decision-Making, Community and Consensual Values: Some Remarks on Braithwaite's Republican Model" [1995] SydLawRw 22; (1995) 17 Syd LR 373; M Krygier and A Glass, "Shaky Premises: Values, Attitudes and the Law" [1995] SydLawRw 23; (1995) 17 Syd LR 385 at 385; E Ghosh, "Republicanism, Community Values and Social Psychology: A Response to Braithwaite's Model of Judicial Deliberation" [1998] SydLawRw 1; (1998) 20 Syd LR 5. However, these critiques do not address Braithwaite's suggestion that judges use republican liberty, but instead focus on Braithwaite's main concern, which is to provide an attractive rationalisation for judges referring to consensual values of the Australian community.

[9] P Pettit, Republicanism: A Theory of Freedom and Government, Oxford University Press (1997). See also his postscript "Republicanism: Once More with Hindsight" for the forthcoming 1999 Oxford University Press paperback edition.

[10] Braithwaite is concerned that in focusing on their justification for punishment, I might convey the impression that they believe that punishment is the major criminal justice issue. Instead, their book seeks to provide a comprehensive approach to the criminal justice system taken as a whole, in which punishment and justifications for punishment are not given too prominent a role. Braithwaite's concern to shift debate away from punishment and towards other approaches to crime is reflected not only in this book, but in many of his other publications, including his Crime, Shame and Reintegration, Cambridge University Press (1989). For one review essay of the latter, followed by Braithwaite's reply and a rejoinder, see C Uggen, "Reintegrating Braithwaite: Shame and Consensus in Criminological Theory" (1993) 18 Law and Social Inquiry 481. My conclusion that republican liberty is unsatisfactory as a sole Value for determining issues relating to punishment leaves untouched many arguments raised in Not Just Deserts and in Braithwaite's other works.

[11] There is some awkwardness in switching between reference to Pettit and reference to Braithwaite and Pettit. While they mention in Not Just Deserts, now 6 supra, p viii, that their book is very much a collaborative project, Pettit bears primary responsibility for the philosophical chapters while Braithwaite is chiefly responsible for the criminological discussion.

[12] [1992] HCA 57; (1992) 177 CLR 292.

[13] Note 1 supra.

[14] W Sadurski in "Book Review" (1991) 10 Law and Philosophy 221 at 223-4 calls it "an excellent book ... I am sure ... [it] will become the focus of discussions on a theory of punishment in years to come". J Kleinig says in his review (1991) 102 Ethics 173 at 175 that they "have approached some traditional questions with a freshness that gives the current discussion a needed charge". A von Hirsch and A Ashworth say in "Not Not Just Deserts: A Response to Braithwaite and Pettit" (1992) 12 O# J of Leg Stud 83 at 84 that the book has the advantage of offering a principled critique of retributivism while other attacks have been atheoretical. A revised version of this review constitutes ch 3 of A von Hirsch, Censure and Sanctions, Oxford University Press (1993). All these reviewers have previously proposed retributivist theories.

[15] Sadurski, note 14 supra at 230-1.

[16] Kleinig, note 14 supra at 174.

[17] Von Hirsch and Ashworth, note 14 supra.

[18] P Pettit, "Republican Themes" (1992) 6(2) Legislative Studies 29; and note 9 supra, p 20.

[19] Pettit, note 9 supra, p 17, referring to I Berlin, Two Concepts of Liberty, Oxford University Press (1958).

[20] Michelman (1986), note 1 supra at 26. Michelman is adopting a Kantian conception of freedom, but states that: "Kant was directly linked to republicanism through Rousseau": note 1 supra at 26.

[21] Ibid at 27.

[22] Pettit, note 5 supra at 163; and note 9 supra, p 19.

[23] Note 5 supra at 164-5; note 9 supra, pp 18-19; and P Pettit, "Liberal/Communitarian: Maclntyre's Mesmeric Dichotomy" in J Horton and S Mendus (eds), After Maclntyre, University of Notre Dame Press (1994) 176 at 188-90.

[24] Now 9 supra, p 18. This differs with his earlier position, where republican liberty was a negative form of liberty albeit one which contrasted with the negative liberty of liberalism: now 6 supra ch 5. He referred then to republican liberty as "resilient non-interference", while he now tends to prefer "non-domination". Nevertheless, the concept remains very largely the same. He now regards the extra republican dimensions of non-interference (legal protection against arbitrary interference and knowledge of that protection) as providing agents with freedom from mastery.

[25] Pettit wishes to be assessed on the basis of how appealing this value is; he says that the historical aspect of his book is secondary: note 9 supra, p 11. For criticism of some aspects of Pettit's depiction of liberalism, see J Levy, "The Liberal Defence of Democracy: A Critique of Pettit" (1994) 29 Australian Journal of Political Science 582. This is followed by P Pettit, "Two Defences of Democratic Voting: A Reply to Levy" (1994) 29 Australian Journal of Political Science 587.

[26] Note 5 supra at 170-1.

[27] Ibid at 166.

[28] Note 9 supra, p 36.

[29] Ibid, pp 45-50.

[30] Note 5 supra at 182-3.

[31] Note 9 supra, p 22.

[32] Ibid, pp 60-1, 71.

[33] Ibid, p 58.

[34] Ibid, pp 148-9. While Pettit refers to taxation as enabling the government to enhance our intensity of liberty, government services also surely provide us with options which extend our liberty. I note, for the sake of completeness, that the extent of freedom can also be reduced by non-interfering factors, that is factors which do not involve an agent intentionally or negligently reducing one's choices. Sickness could be an example.

[35] Ibid, pp 103-6. He also provides other reasons in favour of this priority.

[36] The reasons given by Pettit as to why an egalitarian distribution is required have changed. Previously, he argued that increases in prospects of liberty for those who are best-off do not increase their liberty: P Pettit, "The Freedom of the City: A Republican Ideal" in A Hamlin and P Pettit (eds), The Good Polity: Normative Analysis of the State, Basil Blackwell (1989) 141 at 143. This was rather odd: it was difficult to conceptualise prospects of liberty increasing without liberty itself increasing, since Pettit did not provide us with a sharp distinction between these concepts. An additional complication was that, while Pettit said that we could not consider degrees of perfect liberty within a community, we could consider degrees of perfect liberty in comparing different communities.

[37] Note 9 supra, p 115.

[38] Ibid, p 113.

[39] Ibid, p 117. While I do not subsequently draw upon this aspect of Pettit's discussion, I should point out that according to Pettit the extent of freedom is not necessarily optimised through an egalitarian distribution. This is partly because it is not subject to diminishing marginal productivity. While the utility we derive from additional money tends to diminish as we become wealthier, our choices can be extended at a non-diminishing rate: ibid, p 119.

[40] There may also be readers who accept consequentialism but are sceptical about attempts to define a consequentialist target. However, I suggest in Part IV C that such attempts are useful in furthering discussion of moral questions, but I do not propose a consequentialist target which obviates the need to consider competing perspectives.

[41] See S Scheffler, The Rejection of Consequentialism, Oxford University Press (Revised edition, 1994) for discussion of consequentialism and deontology. See also P Pettit, "The Consequentialist Perspective" in M Baron, P Pettit and M Slote, Three Methods of Ethics, Blackwell Publishers (1997) 92.

[42] I am assuming that the deaths all involve the violation of an identical right. This may not be the case if the agent kills his or her child to avoid two other children being killed by someone who is a stranger to those two children. Deontologists are keen to show that in few cases are we dealing with identical rights: G Grisez and R Shaw, "Persons, Means, and Ends" in J Haber (ed), Absolutism and its Consequentialist Critics, Bowman & Littlefield (1994) 21.

[43] J Haber, "Introduction" in Haber, ibid 1.

[44] They also defend deontology against the charge of paradox: see P Foot, "Utilitarianism and the Virtues" and S Scheffler, "Agent-Centred Restrictions, Rationality, and the Virtues" (which, I think, offers a convincing reply to Foot), both in S Scheffler (ed), Consequentialism and its Critics, Oxford University Press (1988) 224 and 243 respectively. C McMahon in 'The Paradox of Ueontology" (1991) 20 Philosophy & Public Affairs 350 argues that deontological constraints are consistent with maximising a particular value, "interactive fairness" but he recognises that this is not the only value we may wish to maximise.

[45] R Hare, "Ethical theory and utilitarianism" in A Sen and B Williams (eds), Utilitarianism and beyond, Cambridge University Press (1982) 23.

[46] Scheffler, note 41 supra ch 3, drawing on B Williams, "A critique of utilitarianism" in J Smart and B Williams, Utilitarianism For and Against, Cambridge University Press (1973) 77 at 116-17. There are also other approaches to this problem, eg, Pettit, note 41 supra at 163 and M Slote, "Virtue Ethics" in Baron et al, note 41 supra 191.

[47] Scheffler, note 41 supra, p 61.

[48] Of course, there are some commitments which agents should not be allowed to pursue at all and there is also considerable vagueness in the extent to which agents should be allowed to pursue other commitments. I should point out that Scheffler does not concede that consequentialism must violate integrity: ibid, pp 52-5. However, I need not pursue this subtlety here.

[49] On the other hand, one might accept Scheffler's prerogative without accepting his rationale. The prerogative could be regarded as a way of recognising the strength of deontological and consequentialist intuitions, which we may be forced to choose between in certain situations. C Lannore in Patterns of moral complexity, Cambridge University Press (1987) p 148 suggests an alternative way of recognising both intuitions.

[50] There are, in fact, different understandings of utility (see G Scarce, Utilitarianism, Routledge (1996) ch 6) but I need not discuss this since my argument does not involve a comparative assessment of utility versus republican liberty as consequentialist targets. Instead, I argue that on the victimisation issue, republican liberty does not do well enough.

[51] There have been attempts to avoid the conclusion that utilitarianism sanctions repugnant trade-offs between welfare and rights: P Pettit, "The consequentialist can recognise rights" (1988) 38 Philosophical Quarterly 42 fn 1. However, there are doubts about how successful these attempts have been, so it is reasonable for Pettit to suggest an alternative approach.

[52] Braithwaite and Pettit, note 6 supra, pp 37-8.

[53] Ibid, pp 39-40.

[54] This is well set out by Pettit and Braithwaite, ibid, pp 1-7, and I draw upon this.

[55] Ibid, p 2.

[56] See also von Hirsch (1993), note 14 supra, pp 88-91; and A von Hirsch, Past or Future Crimes, Rutgers University Press (1985) pp 3-16.

[57] Note 6 supra, pp 6-7.

[58] This hypothetical, while not taken from L Hinman, Ethics: A pluralist approach to moral theory, Harcourt Brace College Publishers (1994) pp 170-1, is similar to his example.

[59] This may seem a rather crude utilitarian approach: see now 50 supra.

[60] Pettit, note 36 supra at 152; and note 41 supra at 154-5.

[61] Note 51 supra at 53.

[62] Pettit, note 9 supra, p 208.

[63] There are degrees of suspicion between ignorance and knowledge of fabrication. However, incorporating this additional complexity would not change the basic argument.

[64] Mr Smith would suffer substantive arbitrariness if the decision to imprison him reflects a failure to properly take into account his interests: Pettit, note 9 supra, p 55. However, we are trying to determine whether imprisoning Mr Smith is an improper decision, so relying on substantive arbitrariness would involve some circularity. In making a trade-off between Mr Smith's liberty and the liberty of the community, it is relevant whether Mr Smith's liberty is reduced in intensity or extent, yet this is itself dependent upon whether imprisonment is justified. We can avoid these problems by relying upon procedural arbitrariness.

[65] Note 9 supra, p 60.

[66] Braithwaite and Pettit, note 6 supra, p 66 fn 5.

[67] Briefly indicating how they use these aims, incapacitation provides the main justification for imprisonment as opposed to or forms of sentencing. However, imprisonment should only occur in exceptional circumstances. Some punishment is necessary to deter crime. Rehabilitation does not justify punishment, but favours certain types of sentences over others: ibid, pp 124-6.

[68] Ibid, p 127: "We refrain from crime not so much because we fear or even know the punishment we are likely to get [ie deterrence], but because it simply seems wrong to us; and one reason it seems wrong to us is that people are punished and shamed for it".

[69] Ibid, p 88.

[70] Ibid, p 46.

[71] Ibid, p 47.

[72] For empirical evidence of consensus on the relative seriousness of offences, see Braithwaite and Pettit, ibid, pp 178-9; and von Hirsch (1993), note 14 supra, p 29 fn 2.

[73] See text accompanying note 56 supra.

[74] Note 6 supra, pp 176-7.

[75] See also von Hirsch and Ashworth (1992), note 14 supra at 88-9. Braithwaite and Pettit argue further that community concern over disproportionate penalties can be allayed through symbolic politics: murder has a higher maximum penalty than burglary, indicating that the former is more serious. This is surely unrealistic, given the attention the media pay to actual sentences. See also A von Hirsch and A Ashworth "`Dominion' and Censure" in von Hirsch (1993), note 14 supra 20 at 26-7.

[76] Note 6 supra, p 178.

[77] Attribution theorists have argued that we understate the role in controlling behaviour of situational factors and overstate the role of dispositional factors: F Schoeman, "Statistical Norms and Moral Attributions" in F Schoeman (ed), Responsibility, Character, and the Emotions, Cambridge University Press (1987) 287.

[78] For further arguments concerning the politics of relying on desert-based approaches, see von Hirsch (1993), note 14 supra, ch 10.

[79] lbid, ch 8. One rationale is that punishment cancels the unfair advantage a criminal enjoys by violating society's system of mutual restraints: M Davis, "Criminal desert and unfair advantage: What is the connection?" (1993) 12 Law and Philosophy 133. Another rationale relies on the expressive function of punishment, that is it denounces crime: von Hirsch (1993), note 14 supra.

[80] Note 6 supra, p 34. Negative retributivism imposes a ceiling upon punishment: it cannot be greater than the punishment which is proportional to culpability. Positive retributivism imposes a floor upon punishment: punishment must not be less than the degree commensurate with the nature of the crime and the culpability of the criminal. Intrinsic retributivism contrasts with theories which justify retributivism by reference to other values.

[81] With respect to defending positive retributivist principles on the ground of their intrinsic value, see M Moore, "The Moral Worth of Retribution" in Schoeman (ed), note 77 supra 179.

[82] Note 6 supra, p 207.

[83] Instead, they argue that the forms of retributivism referred to in note 80 supra have a consequentialist flavour: note 6 supra, pp 48-52.

[84] This situation can be imagined with the following hypothetical. Assume that Mahatma Gandhi's assassin was a Muslim, not a Hindu. Framing a Hindu, assuming that this could be accomplished with no risk of detection, may be justifiable as a means of avoiding many Muslims being killed in reprisals, followed by counter-reprisals against Hindus etc. A deontologist might endorse framing the Hindu if the consequences of refraining from doing so are regarded as disastrous. However, the difficulty with this approach is that the deontologist might look like a covert consequentialist, prepared to abandon agent-centred constraints when they are consequentially unjustifiable.

[85] For an approach which considers the wider consequences of punishment in terms of desert, see T Clear, "The punitive paradox: desert and the compulsion to punish" (1996) 33 Journal of Research in Crime and Delinquency 94. See W Sadurski, Giving Desert its Due, D Heide] Publishing Company (1985); and G Sher, Desert, Princeton University Press (1987) for discussions of desert applied to a range of issues, including punishment.

[86] In terms of official policy, there are good consequentialist grounds for limiting the authority of government agents so that they do not take into account such adverse consequences of following legal procedures. However, the more difficult question is whether the agent should respect this constraint. It may seem odd, in the light of my discussion of the hypothetical detective, to leave it on for agents to exceed their authority. However, my argument is not that desert must always lead us to respect legal procedures: see Part III D.

[87] The fundamental distinction between consequentialism and deontology lies over the former's use of an impersonal standpoint while the latter insists on agent-centred constraints: see Part H C. However, due to utilitarianism often being understood as the consequentialist theory, theories differing in significant respects from utilitarianism have often been classed as non-consequentialist without proper consideration of whether this is appropriate. For example, a theory that suggests certain principles or values as constraints upon outcomes, or involves a value which is past- rather than future-oriented (eg desert), need not be understood deontologically.

[88] The idea of lexical ranking was suggested in T Scanlon, "Rights, Goals and Fairness" in S Hampshire (ed), Public and Private Morality, Cambridge University Press (1978) 93. For an example of a theory employing lexical ranking, see D Cummiskey, Kantian Consequentialism, Oxford University Press (1996).

[89] J Rawls, A Theory of Justice, Oxford University Press (1972) p 75.

[90] Scheffler, note 41 supra, p 31.

[91] Retributivist theories tend to be deontological. Von Hirsch, for example, may be concerned that I do not exclude the option of punishing burglars more severely than murderers. The deterrent benefits assumed in increasing the penalties for burglary are not simply disregarded on the basis of the constraints imposed by retributivism. Instead, deterring the imposition of undeserved suffering on potential victims of burglary should be considered, along with the injustice of the disproportionate penalty. My argument differs from preventionism or Braithwaite and Pettit's approach by making the injustice of the disproportionate penalty a relevant and weighty consideration.

[92] Davis, note 79 supra.

[93] However, I did touch upon this issue in Part III B. I think that positive retributivist principles should be given substantial weight. However, this does not mean that they cannot be compromised in order to further other values. A general preoccupation of Braithwaite's has been to demonstrate the benefits which derive from not following positive retributivist principles. The extent to which other values should be permitted to compromise these principles is discussed in von Hirsch (1993), now 14 supra, ch 6. new is also discussion of hybrid theories from a perspective opposed to desert in N Walker, Why Punish?, Oxford University Press (1991) pt IV.

[94] Note 6 supra, pp 39-40.

[95] As Braithwaite and Pettit say, "no theory which countenanced something like the victimization of the innocent could be taken seriously": note 6 supra, p 43. A theory must provide a stable allocation of the rights which are uncontroversial in the relevant community: ibid, pp 42-3.

[96] In note 51 supra at 54, Pettit admits that if dissemblance is possible (ie calculating while convincing rights-holders that one is not doing so), this may seem optimal. However, he suggests that this strategy is not available to the state since its deliberations are a matter of public record.

[97] Note 12 supra.

[98] The majority consisted of Mason CJ, Deane, Toohey, Gaudron and McHugh JJ. Brennan and Dawson JJ dissented. The limitations of the entitlement found by the Court are set out in G Zdenkowski, "Defending the indigent accused in serious cases: a legal right to counsel?" (1994) 18 Crim LJ 13

[99] Note 8 supra and accompanying text. I have omitted a step in Braithwaite's approach. He suggests that judges refer to consensual community values in hard cases. When community values conflict in their application to a case, he suggests that republican liberty be used as a yardstick to balance these conflicting values. However, Braithwaite gives judges considerable freedom in interpreting community values so it is reasonable to assume that community values may well come into conflict in hard cases. Considering community values in terns of their contribution to republican liberty seems roughly equivalent to deciding cases on the basis of which outcome maximises republican liberty.

[100] Note 6 supra, p 43.

[101] Mason CJ and McHugh J stated in their joint judgment that: "the common law of Australia does not recognise the right of an accused to be provided with counsel at public expense": note 12 supra at 297-8. Instead, "Australian law acknowledges that an accused has the nght to a fair trial and that, depending on all the circumstances of the particular case, lack of representation may mean that an accused is unable to receive, or did not receive, a fair trial": at 311. These statements reflect the majority view of the Court. The Court did not declare a right to the provision of legal representation because the entitlement it found was only indirectly enforceable (through court stays) and was qualified, not absolute. Interestingly, the Court did not consistently apply this narrow understanding of rights. Mason CJ and McHugh J, citing Deane J in Jago v District Court (NSW) [1989] HCA 46; (1988) 168 CLR 23 at 56-7, said that strictly speaking there is no right to a fair trial for no person can enforce a nght to be tried by the state. Instead, there is a right not to be tried unfairly. However, "it is convenient, and not unduly misleading, to refer to an accused's positive right to a fair trial": at 299. There is no reason to understand Braithwaite and Pettit to be restricting `rights' to absolute entitlements directly enforceable against the state. Braithwaite and Pettit's theoretical support for rights can therefore be tested by reference to Dietrich and it is convenient in this context to refer to Dietrich as establishing a "right to the provision of legal representation".

[102] Braithwaite, note 7 supra; and Pettit, note 9 supra, ch 6.

[103] Dietrich was in fact acquitted of the fourth count: note 12 supra at 298 per Mason CJ and McHugh J, and at 365-6 per Gaudron J.

[104] A brief account of Commonwealth involvement in legal aid, albeit one which predates recent Commonwealth initiatives to withdraw from funding of State matters, is given in Access to Justice Advisory Committee, Access to Justice: an Action Plan, Australian Government Publishing Service (1994) pp 228-31. See also C Steytler, "Legal aid cuts" (1997) 71 ALJ 89. The prospect of criminals going free has been explicitly raised in the media. The NSW Attorney-General has referred to this possibility, in relation to criminals who faced a hung jury in their initial trial: B Lagan and J Brough, "States to hit back on legal aid cuts" Sydney Morning Herald, 7 February 1997, p 3; and G Zdenkowski, "Legal aid is crucial" Sydney Morning Herald, 7 February 1997, p 15.

[105] Note 12 supra at 312.

[106] Ibid. See also Zdenkowski, note 98 supra at 151-2.

[107] Note 6 supra, p 87.

[108] Michelman (1988), note 1 supra at 1503, 1507.

[109] Michelman (1988), note 1 supra at 1532, 1537.

[110] R Dworkin, Taking Rights Seriously, Duckworth (1977) pp 90-100.

[111] For criticisms of Dworkin's suggestion that judges only apply principle and avoid utilitarian policy considerations, see J Bell, Policy Arguments in Judicial Decisions, Clarendon Press (1983) ch 8.

[112] Pettit contrasts pluralism with republicanism in note 9 supra, p 202.

[113] Note 12 supra at 349-50. Perhaps more precisely, it could be said that justice for unrepresented people charged with serious criminal offences cannot be pursued in isolation for there are competing demands upon the public purse, including demands based on justice.

[114] Ibid at 365.

[115] Ibid at 372.

[116] Ibid at 312.

[117] Interestingly, in elaborating a theory of judicial deliberation, Braithwaite refers approvingly to Michelman's model: note 7 supra at 368 fn 59. W Rich in "Approaches to Constitutional Interpretation in Australia: an American Perspective" (1993) U Tas LR 150 has also favoured a republican approach which is similar to Michelman's.

[118] Michelman (1986), note 1 supra at 31.

[119] Michelman (1988), note 1 supra at 1512 fn 70. Scepticism about comprehensive theories is also found in the work of another important republican legal scholar: C Sunstein, Legal Reasoning and Political Conflict, Oxford University Press (1996) p 43. Returning to Michelman, he is also non-committal between deontological and consequentialist approaches: F Michelman, "Always under Law?" (1995) 12 Constitutional Commentary 227 at 232 fn 1 I. It might be argued that my proposal to place desert and other values within a consequentialist framework is inconsistent with Michelman's approach. However, articulating a consequentialist target does not necessarily exclude deontological intuitions: see note 49 supra and accompanying text. Also, my proposal is made in the context of critiquing Pettit; I suggest what seems to me a more attractive target. I think it should be considered in republican deliberation, but it is not intended to preclude consideration of deontological intuitions.

[120] Note 6 supra, chs 3, 4.

[121] The prosecution and judge have some discretion as to the extent to which they take into account the vulnerable status of an unrepresented accused. It is of little comfort to those unrepresented that this discretion is confined by the adversarial nature of our system. For example, judges and perhaps even tribunal members are limited in the assistance they can provide by the principle of non-intervention: M Allays, "Neutrality, the Judicial Paradigm and Tribunal Procedure" [1991] SydLawRw 26; (1991) 13 Syd LR 377 at 385-91 and 406-11. With respect to tribunals, cf M Allays, Administrative Law: Cases and Materials, Butterworths (1997) pp 382-3. The limitations of the forms of assistance in ensuring a fair trial were recognised in Dietrich: note 12 supra at 302 per Mason CJ and McHugh J, at 335 per Deane J, and at 369-71 per Gaudron J.

[122] Michelman's model of deliberation, being based on a broad political ideal, is also relevant to politicians and citizens in general. Without their support, the status of Dietrich may be precarious since only Deane and Gaudron JJ held that the right to a fair trial was constitutionally entrenched: J Hope, "A constitutional right to a fair trial? Implications for the reform of the Australian criminal justice system" (1996) 24 Fed LR 173. I note that the Commonwealth Attorney-General was reported to be planning to ask States to override or wind back the Dietrich decision: M Kingston, "Legal aid for the poor under threat" Sydney Morning Herald, 28 March 1998, p 8.

[123] Note 6 supra, p 126, referring to Braithwaite, note 10 supra.

[124] Sadurski, note 14 supra at 230-1.

[125] Note 9 supra, p 81.

[126] Ibid, p 3.

[127] Section 2 of the postscript, note 9 supra.

[128] Note 9 supra, p 12.

[129] Ibid, p 137.

[130] I am not implying, of course, that his book loses all justification if republican liberty is unsatisfactory as the sole goal for our political system. I agree with Pettit that his interpretation of republicanism can still retain its interest: see section 2 of the postscript, note 9 supra.

[131] For example, Pettit's suggestion that environmental concerns be voiced through the language of republican liberty leads to the exclusion of ecocentric values, yet it accommodates the argument that damaging the environment reduces our opportunity to affirm "our conaturality with other species": note 9 supra, pp 136-7. Can ecocentric values be excluded without significant loss? Does affirming "our conaturality with other species" provide a significantly less contentious or complex reason for protecting the environment?


AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback
URL: http://www.austlii.edu.au/au/journals/UNSWLawJl/1999/43.html