Suspended sentences and preventive sentences: illusory evils and disproportionate punishments mirko bagaric

A suspended sentence threatens future harm for criminal conduct that has already occurred. It is a term of imprisonment, the execution of which is wholly or partly suspended. Ostensibly, it is a heavy sanction. Available in most Australian jurisdictions and in the United Kingdom, it is frequently employed to punish serious breaches of the criminal law. For example, in 1996 there were over five thousand suspended sentences imposed in Victoria alone.¹ During this period, in the County and Supreme Courts of Victoria (which have jurisdiction over the most serious criminal offences) the suspended sentence was the second most commonly imposed sanction, comprising about 30 per cent of all sanctions.² On the face of it, such figures are unremarkable and are unlikely to prompt consternation, since the popularity of the suspended sentence among sentencers is matched by the enthusiasm for them among recipients. It has been noted that “a defendant who has committed an offence so serious as to merit imprisonment but who has had that sentence wrongly suspended is obviously more likely to be out celebrating than dashing to the Court of Appeal”.³ It will be argued that there is good reason for offenders’ enthusiasm towards suspended sentences; they do not constitute a recognisable form of punishment at all.

This paper considers the nature of the suspended sentence, particularly the pragmatic and conceptual difficulties with it. Following an analysis of the concept of punishment, it is contended that the suspended sentence is merely an illusory unpleasantness and should therefore be abolished as a sentencing option. Further, recent recommendations to reintroduce suspended sentences as a sentencing option in New South Wales,⁴ the only jurisdiction in Australia where it is presently unavailable, should be rejected.

The suspended sentence will then be compared with the preventive (or protective)⁵ sentence, which, it is argued, is the logical converse of the suspended sentence. A preventive sentence inflicts immediate harm on an ‘offender’, normally in the form of imprisonment, on account of threatened future criminal conduct; while the suspended sentence threatens a future evil (in the form of restoration of the term of imprisonment which has been suspended) for criminal conduct that has already occurred. Suspended and preventive sentences are also alike in that both violate the principle of proportionality, which forms another basis on which suspended sentences should be abolished. Despite this symmetry, suspended sentences are generally widely accepted, while preventive sentences are almost universally condemned.

Sanctions in the form of suspended sentences have a long history, their first use can be traced back to the ecclesiastical courts in the fourteenth century.⁶ Today, the suspended sentence is available as a sentencing option in the United Kingdom and all Australian jurisdictions,⁷ except New South Wales where it was abolished in 1974.⁸ Suspended sentences have been subject to the greatest amount of empirical analysis in Victoria and England, and hence this discussion will focus largely on their availability and use in these jurisdictions. They were introduced in Victoria in 1915,⁹ but were not available in England until 1967.

In these jurisdictions, suspended sentences are regarded as heavy sanctions. For example, in Victoria they rank fourth in the hierarchy of gravity of sanctions behind immediate terms of imprisonment, combined custody and treatment orders,¹⁰ and intensive corrections orders.¹¹ They are commonly described as a threat perched like the Sword of Damocles over the head of offenders during the period of operation.¹² All terms of imprisonment of not more than three years may be wholly or partly suspended in Victoria,¹³ and the maximum operational period of a suspended sentence, the period during which the offender must not commit another offence, is three years.¹⁴ The position is similar in England, where any sentence of imprisonment of two years or less may be wholly or partly suspended for a period of between one and two years.¹⁵ The reason that suspension is allowed only in relation to relatively short sentences of imprisonment is because it is felt that any sentence beyond this would be for an offence that is so serious that it would be inappropriate to suspend punishment.¹⁶

Where a suspended sentence is breached there is a presumption favouring its restoration.¹⁷ In the United Kingdom and Victoria, if the offender commits an imprisonable offence during the period of the suspended sentence the court must activate the suspended sentence and commit the offender to prison, unless it would be unjust to do so.¹⁸ In Victoria, the presumption is even stronger because in determining if it would be unjust to activate the term of imprisonment only exceptional circumstances may be considered.¹⁹

On balance, suspended sentences are viewed favourably by courts and commentators. However, they have come under criticism in two respects. First, on the basis that the reasoning process leading to their imposition is logically unsound. Secondly, that they have been unsuccessful in achieving their (perceived) aim of reducing prison numbers. Although both these criticisms are of some merit, I will argue that neither constitutes a decisive attack on suspended sentences as a sentencing option. I will then discuss what I consider to be a far more persuasive objection to suspended sentences.
III. Criticisms of Suspended Sentences

A. The Reasoning Process Underlying Imposition of Suspended Sentences

(i) Conceptual Incongruity Underlying the Imposition of Suspended Sentences

A paradoxical aspect of suspended sentences is that, strictly speaking, they may be imposed only where it is felt that an immediate custodial sanction is appropriate.²⁰ The court must first reach the conclusion that an immediate term of imprisonment is warranted, fix the sentence and only then consider whether to suspend the sentence.²¹ The absurdity in such an approach stems from the fact that an immediate term of imprisonment is a sanction of last resort; it can only be imposed if the sentencer is satisfied that the purpose or purposes for which the sentence is imposed cannot be achieved by a sentence that does not involve the confinement of the offender.²² If all of the factors in mitigation have been considered at the outset and an immediate custodial sentence is imposed, there is nothing left which can reduce the severity of the penalty.²³ Once sentences higher up in the sentencing hierarchy than a suspended sentence have been dismissed as too mild, it is farcical to claim that a suspended sentence is appropriate, particularly when there are no new variables to tip the scales further in favour of a more lenient disposition. It is an affront to both the laws of physics and logic to propose that vacuity can produce change.

However, the main purpose in suspending a sentence is to encourage reform of the offender,²⁴ and thus the main consideration in determining whether or not to suspend a sentence is the prospects of rehabilitation.²⁵ This accords with the historical aim of the suspended sentence which is to prevent criminal behaviour, rather than to match a penalty with the gravity of the offence.²⁶ In view of this, the reasoning process behind suspended sentences can be defended by arguing that while such sentences are only imposed where it is determined that an immediate sentence is appropriate, a softening in the sanction can occur where the offender has particularly good prospects for rehabilitation. However, this is unsound. Prospects of rehabilitation are, and should be, factored into the initial sentencing determination, rather than counted twice.

The confusion that the above approach encourages is illustrated by the comments of an English Magistrate, JQ Campbell, who in opposition to the changes placing stricter limits on the availability of suspended sentences in the United Kingdom since October 1992,²⁷ stated that “if I am dealing with a case where I would have suspended a custodial sentence prior to October 1992 but now feel prevented from doing so it would be fundamentally unjust to impose an immediate custodial sentence”.²⁸ This sentiment is clearly erroneous. A suspended sentence should not have been imposed in the first place if it was unjust to impose an immediate term.

In light of the incongruity of the logical reasoning underlying the imposition of suspended sentences, it is hardly surprising that certain anomalies or unanticipated consequences have emerged regarding their use. Empirical studies reveal that only about half of suspended sentences imposed appear to represent a diversion from immediate custodial sentences, while the other half reflect net widening, that is, imposing a suspended sentence in circumstances where a less severe penalty would otherwise have been imposed.²⁹ Suspended sentences have also resulted in a trend towards sentence inflation, whereby offenders are given extra time in light of the term being suspended. A survey by Tait regarding the use of suspended sentences in Victoria during the period 1985-91 showed that for Victorian magistrates the inflation rate was about 50 per cent.³⁰

In terms of the main recipients of suspended sentences, evidence seems to support the view that they are used largely as a means of appearing tough on those who are normally treated leniently anyway: middle-class offenders and those with a settled life style.³¹ A survey by Moxon in 1988 of Crown Courts disclosed that suspended sentences were common in breach of trust of cases, typically involving white collar workers.³² Where they were imposed on those with a criminal record, this was generally in relation to those who appeared to have a more settled future.³³

The absurdity associated with the reasoning process behind suspended sentences is not, however, a persuasive reason for their abolishment. For this is merely a contingent matter which has no bearing on the intrinsic character of the suspended sentence as a criminal sanction. For example, a necessary and sufficient precondition to a suspended sentence could just as easily be that it is the most appropriate sanction in light of its ranking in the sentencing hierarchy, whatever this might be. However, if suspended sentences are to remain a viable sentencing option the need for transparency and intellectual honesty requires revision regarding the circumstances in which they may be imposed.

In England, the suspended sentence was introduced without detailed consideration of its use in other countries³⁴ as part of an effort to reduce prison numbers.³⁵ To this end it appears to have failed:

the accumulated evidence is not encouraging. If the main object of the suspended sentence was to reduce the prison population, there are considerable doubts as to whether it has achieved this effect. It may have even increased the size of the prison population.³⁶

It is suggested that there were three reasons for this failure. First, on many occasions accused received suspended sentences where previously they would have received a non-custodial order such as a fine. Secondly, the term of a suspended sentence was generally longer than an immediate custodial sentence and upon breach the term was often implemented in full and consecutively. Finally, for the next offence committed after a suspended sentence, the natural penalty was a period of imprisonment.³⁷

Other reasons advanced for the failure of the suspended sentence are that:

[It was seen as a] convenient via medium, midway between the custodial and non-custodial penalties, so that courts previously hesitating between the two and coming down on the side of non-custodial penalties would now choose the suspended sentence as an obvious alternative; and secondly, that many [sentencers] did not share the official Government thinking behind the introduction of the suspended sentence, and saw it not as an alternative to prison but as an especially effective Sword of Damocles which would deter individual offenders much more surely than probation or the fine.³⁸

Despite the apparent failure of suspended sentences to live up to expectations, the Advisory Council on the Penal System on Sentences of Imprisonment in its report in 1978 proposed no change in relation to suspended sentences.³⁹ It stated that one of the benefits of the suspended sentence was that it provided courts with a sanction allowing offenders to avoid actual imprisonment; “roughly three quarters of offenders given suspended sentences are not imprisoned for the offence for which the suspended sentences were given”.⁴⁰ However, as Bottoms has pointed out, “it is the ultimate impact on the prison population of the whole effect of a suspended sentence, not just the apparent immediate impact, which really matters for penal analysis”.⁴¹

The ineffectiveness of the suspended sentence in reducing the prison population is demonstrated by the fact that since it was abolished in England in 1982 for offenders under the age of 21, there is no evidence that this resulted in an increase in the number of immediate custodial sentences regarding such offenders.⁴²

The dissatisfaction with the suspended sentence as a punitive measure in England culminated with measures being taken to reduce its use. As a result of changes introduced by the Criminal Justice Act 1991 (UK), which came into effect in October 1992, the use of suspended sentences, in terms of the overall number of penalties imposed, fell from ten per cent to one per cent for males and from eight per cent to two per cent for females.⁴³ This change occurred primarily because it became a requirement that custodial sentences were to be suspended only in exceptional circumstances.⁴⁴ Obviously, the message that the suspended sentence “should be used far more sparingly than it has been in the past”,⁴⁵ was clearly received by the courts.
(ii) The Experience in Victoria

The English experience of suspended sentences is in contrast to that in Victoria, where the overall impact of suspended sentences has resulted in a reduction in the prison population.⁴⁶ As I adverted to earlier, the suspended sentence is a widely utilised sanction in Victoria. For example, in 1991 it accounted for five per cent of all sanctions imposed in the Magistrates’ Court and 20 per cent of sanctions imposed by the County and Supreme Court (the Higher Courts).⁴⁷ By 1996 this had grown to about six per cent and 30 per cent respectively.⁴⁸

In Victoria the breach rate for suspended sentences in 1990 was 18 per cent, which was less than half the rate in England. The activation rate for those breaching suspended sentences was also significantly less in Victoria than in England: 54 per cent, compared to about 80 per cent.⁴⁹ The difference in the breach rates can be explained on the basis that the length of the operational period of suspended sentences in England was up to three years as opposed to one year in Victoria.⁵⁰

Accordingly, to the extent that their objective is to reduce prison numbers, suspended sentences have succeeded in Victoria. In light of this, Tait concludes that:

[Suspended sentences] are still something of a mystery. They threaten future pain to ensure present compliance. They depend for their success on the avoidance of certain behaviours rather than the performance of activities. They appear to be inconsistent with other forms of penalty which extract money, work, reporting behaviour or loss of liberty. In a system which prides itself on proportionality and consistency, it is hard to make a case for an invisible, intangible, but frequently irresistible sanction. Except that it works.⁵¹
(iii) Evaluating Suspended Sentences by Reference to Reduction in Prison Numbers

However, Tait’s argument is flawed. Sentencing options cannot be evaluated on the basis of their impact on the frequency with which other sentencing options are used. Otherwise it could be argued that mandatory prison sentences for road traffic offences are desirable because they would reduce the amount of fines issued. More particularly, the effect on the prison population is not a weighty, far less the sole, consideration by which the success of a criminal sanction may be assessed. If keeping people out of jail is the measure of success, absolute victory could be achieved by merely opening the prison gates. Less drastically, prisons would be almost totally emptied by converting every prison term of less than twelve months automatically into another sanction such as probation or a fine. But, as should be apparent by now, such suggestions totally miss the point. Indeed in many circumstances it may be that keeping people out of prison is undesirable.

The crucial, and indeed only, question in relation to the effectiveness of sentencing options is whether they fulfil the objectives of a properly considered and coherent system of punishment. Imprisonment is not an objective, but rather a means, of punishment. Unfortunately, a meaningful analysis of the extent to which the suspended sentence promotes the objectives of punishment is not possible, given that a primary sentencing rationale has not been adopted by any sentencing system in Australia or the United Kingdom. This is in keeping with a worldwide phenomenon: the sentencing codes of most countries do not expressly adopt a particular theory of punishment,⁵² and where sentencing objectives are declared they are often inconsistent.⁵³ Good examples are the Sentencing Act 1991 (Vic) and the Criminal Justice Act 1991 (UK).

The Sentencing Act 1991 (Vic)⁵⁴ expressly adopts the utilitarian based goals of deterrence, rehabilitation and incapacitation on the one hand, while simultaneously promoting (apparently inconsistent) retributive objectives such as denunciation and just deserts, and then provides that these five purposes are exhaustive of the purposes for which sentences may be imposed. However, by failing to prioritise the respective importance of these objectives it seems that they were adopted in blissful ignorance of any inconsistency or tension between them. The Criminal Justice Act 1991 (UK) fares no better. The White Paper upon which the Act is based clearly supported a retributive theory of punishment: “the first objective of all sentencing is denunciation and retribution for the crime”.⁵⁵ In light of this it has been suggested that the Act gives desert and proportionality a primary role,⁵⁶ yet nowhere in the Act is this made express and in fact the Act states nothing about the rationales for sentencing. Indeed the only consideration which in certain circumstances can trump all others is incapacitation, which is clearly a utilitarian goal.⁵⁷

The unprincipled nature of sentencing practice has led to what Andrew Ashworth labels a ‘cafeteria system’ of sentencing, whereby sentencers may pick and chose a rationale which seems appropriate at the time with little constraint. This is made significantly easier by the large number of discrete factors that the courts have identified as being relevant to sentencing. Two separate studies about twenty years ago determined that there were between 200 and 300 such factors.⁵⁸

In any event, the failure of legislatures to develop a primary rationale for sentencing is not significant for the purposes of this discussion, since as is discussed below suspended sentences fail at the first hurdle: they do not constitute a form of punishment.

Despite the ostensible severity of suspended sentences, there have been some reservations expressed about their punitive impact. In R v King, Lord Parker stated that “in many cases [where a suspended sentence is imposed] it is quite a good thing to impose a fine, which adds a sting to what might otherwise be thought by the prisoner to be a let-off”.⁵⁹ Even reports generally supportive of suspended sentences as a sentencing option have acknowledged that “excessive mitigation [is] inherent”⁶⁰ in them. The 1990 Home Office White Paper noted that:

Many offenders see a suspended sentence as being a ‘let off’ since it places no restrictions other than the obligation not to offend again. If they complete the sentence satisfactorily, all they have felt is the denunciation of the conviction and sentence, any subsequent publicity and, of course, the impact of acquiring a criminal record.⁶¹

Against this, the suspended sentence has been described as a significant punishment,⁶²which carries a serious stigma.⁶³ Further, suspended sentences have been defended on the basis that the most effective way to prevent criminal behaviour is by internal restraints stemming from education and socialisation and that a threat of punishment is “just as ‘real’ as any of the other fears, expectations, obligations, and duties which populate the social world, ... and this threat is more individualised and immediate when a court imposes such a sentence”.⁶⁴ It has also been suggested that the suspended sentence may be conceived as punishment since it “is not something which the offender welcomes in itself”.⁶⁵

To get to the bottom of whether a suspended sentence constitutes a form a punishment it is necessary to first investigate the essential nature of punishment and then to break down the suspended sentence into its constituent parts to ascertain how it squares with the concept of punishment.
(i) The Nature of Punishment

An enormous number of definitions of punishment have been advanced over the ages. Many incorporate somewhat controversial aspects into the definition, for example, by confining punishment to the guilty.⁶⁶ Thus, Herbert Morris defines punishment as “the imposition upon a person who is believed to be at fault of something commonly believed to be a deprivation where that deprivation is justified by the person’s guilty behaviour”.⁶⁷ Duff defines punishment as “the infliction of suffering on a member of the community who has broken its laws”;⁶⁸ and similarly McTaggart defines punishment as “the infliction of pain on a person because he has done wrong”.⁶⁹

Other definitions include the purported requirement that punishment must be inflicted by an appropriate authority.⁷⁰ For example, Hobbes provides that

a punishment is evill inflicted by ublique Authority, on him that hath done, or omitted that which is judged by the same Authority to be a transgression of the law; to the end that the will of men may thereby the better be disposed to obedience... The aym of punishment is not revenge, but terrour.⁷¹

Honderich defines punishment as “an authority’s infliction of a penalty, something involving deprivation or distress, on an offender, someone found to have broken a rule, for an offence, an act of the kind prohibited by the rule”.⁷² And in the postscript to the same book, written over a decade later, as “that practice whereby a social authority visits penalties on offenders, one of its deliberate aims being to do so”.⁷³

For the purposes of the present discussion it is not necessary to resolve timeless disputes about matters such as by whom punishment must be imposed and whether or not punishment is confined to the guilty. For present purposes, what is telling is that an indispensable feature of any tenable definition of punishment is that it must constitute some inconvenience to the offender.

Thus Bentham simply declared that “all punishment is mischief, all punishment is itself evil”.⁷⁴ Ten states that punishment “involves the infliction of some unpleasantness on the offender or it deprives the offender of something valued”.⁷⁵ Others have placed somewhat emotive emphasis on the hurt that punishment seeks to secure. Punishment has been described as simply pain delivery,⁷⁶ and similarly it has been asserted that “the intrinsic point of punishment is that it should hurt - that it should inflict suffering, hardship or burdens”.⁷⁷ Walker is somewhat more expansive regarding the type of evils which can constitute punishment: he describes punishment as

the infliction of something which is assumed to be unwelcome to the recipient: the inconvenience of a disqualification, the hardship of incarceration, the suffering of a flogging, exclusion from the country or community, or in extreme cases death.⁷⁸

Finally, von Hirsch states that “punishing someone consists of doing something painful or unpleasant to him, because he has purportedly committed a wrong, under circumstances and in a manner that conveys disapprobation of the offender for his wrong”.⁷⁹

The requisite inconvenience flowing from punishment has been described in numerous ways, including evil, pain, suffering, or hurt. Linguistic creativity aside, the important point to emerge is that despite continuing unresolved issues about the nature of punishment, one settled feature is that punishment involves an unpleasantness imposed on the offender.⁸⁰ This incontrovertible and seemingly innocuous truth is fatal to the continuation of the suspended sentence as a sentencing option.

The suspended sentence has two components. The first is the term of imprisonment which is imposed. Clearly, it cannot be argued that this constitutes a form of unpleasantness since by the very nature of the sanction it is suspended precisely in order to avoid its effective operation. The other aspect of the suspended sentence is the possibility that the period of imprisonment may be activated if a condition related to the sentence, namely that the offender not re-offend, is breached during its operation.⁸¹ And it is this feature of the suspended sentence which supposedly carries the sting. Accordingly, although the suspended sentence contains no tangible inherent unpleasantness, a real unpleasantness is imposed since the people undergoing it face the risk of activation in the event of a breach.

However, it is erroneous to describe such a risk as being capable of comprising a punitive measure. Every person in the community faces the risk of imprisonment if they commit an offence which is punishable by imprisonment.⁸² In this way the natural and pervasive operation of the criminal law casts a permanent Sword of Damocles over all our heads: each action we perform is subject to the criminal law. Despite this it has never been seriously asserted that we are all undergoing some type of criminal punishment. It follows logically that the risk of imprisonment in the event of a future commission of a criminal offence is not a criminal sanction; it is a nullity in terms of punitive effect. The situation is obviously somewhat more precarious for those undergoing suspended sentences: in addition to the risk faced by all of us of imprisonment if we commit a criminal offence, they have the more specific risk that commission of an offence may also result in them being imprisoned by virtue of restoration of the sentence which is suspended. But this additional risk is of precisely the same nature⁸³ (the possibility of imprisonment in the event of committing an offence) as that borne by all members of the community. It is irrelevant that for those undergoing suspended sentences the likely level of unpleasantness is greater should the risk eventuate; the difference is one of degree, not nature. It is important to note that this conclusion follows not from a ‘mere’ value judgement, but is rather an irresistible mathematical truth: two times zero is still zero.

The illusory punitive nature of the suspended sentence is emphasised, in that not only offenders who breach suspended sentences receive a greater penalty than is warranted by the immediate offence. Offenders with prior convictions are also typically dealt with more harshly than those without a criminal record. Though offenders are not punished again for their previous crimes,⁸⁴ the earlier offending may disentitle them from leniency by not allowing a reduction in sentence for good character.⁸⁵ Despite this it cannot be contended that offenders who have ‘served their time’ are still undergoing punishment.

Thus the true picture seems to be that the suspended sentence suffers from the fundamental flaw that it does not constitute a discernible unpleasantness. Rather it merely signifies a possible future unpleasantness: if there is no breach, there is no evil.⁸⁶ Moreover, given that avoidance of the unpleasantness for those undergoing suspended sentences is totally within their control, during the period of ‘sentence’ they are in the identical position as the rest of the community, in so far as being subject to criminal sanctions is concerned. The equation is the same: offend and risk jail; abide by the law and suffer no unpleasantness.
(iii) Community Attitudes Regarding Suspended Sentences

Surveys regarding community attitudes about the ranking of penalties have shown that few are deceived by the superficial punitive veneer of the suspended sentence. A survey conducted in Philadelphia and Pennsylvania asked respondents (consisting of a group of police officers, a group of inmates, a group of probation officers and an undergraduate criminology class) to rank 36 different penalties, ranging from death to a $10 fine in order of severity.⁸⁷ These penalties included suspended sentences of three years, 12 months and six months.⁸⁸ The mean rank orders for these from the four groups⁸⁹ were 27, 30 and 32 respectively. All of the suspended sentences ranked in order of severity below a fine of $500⁹⁰ and above a fine of $250.⁹¹ It was concluded that “a suspended sentence involving the prospect of a possible prison sentence for a specified term is less burdensome than the immediate inconvenience of probation supervision or a financial penalty”.⁹²

In England, a survey revealed that members of the public viewed the suspended sentence as the least punitive sanction of seven common penalties. The suspended sentence was considered more lenient than probation and even softer than a small fine.⁹³ Such a view appears to be widespread. A survey of lay justices found that suspended sentences of six months were regarded as more lenient than probation of two years and both of these sanctions were below a one hundred pound fine.⁹⁴ The conclusion to be drawn is that suspended sentences are regarded as more lenient than almost any sentence of peremptory punishment, and “although apparently second only to immediate imprisonment on the sentencing hierarchy, [the suspended sentence] is treated in practice as an option much lower down the ladder”.⁹⁵

The inadequacy of suspended sentences as a punitive measure is further illustrated by comparing them with their converse: protective (or preventive) sentences.

A preventive sentence is a sanction that is imposed in response to some future harm that it is anticipated the ‘offender’ may commit. Morris neatly encapsulates the essence of preventive sentences by comparing them to pre-emptive strikes: “in the criminal law, if not in international relations, the pre-emptive strike has great attraction; to capture the criminal before the crime is committed is surely an alluring idea”.⁹⁶

Thus the protective sentence imposes a present evil, normally in the form of imprisonment, for criminal behaviour which has not as yet occurred and may in fact never occur.⁹⁷ It is aimed at people whose perceived propensity for engaging in violent⁹⁸ behaviour is so high that they are an unacceptable risk to the community.

Two other types of sentences have also loosely been referred to as protective sentences: indefinite sentences and additional fixed sentences.⁹⁹ Indefinite sentences are penalties imposed without a termination date. They can be imposed at the outset or as an extension of a normal sentence. Indefinite sentences are typically reviewable at defined intervals by a court and are now available in many Australian jurisdictions.¹⁰⁰ Additional sentences are sanctions that are imposed beyond that which is appropriate for the particular offence, normally due to previous offences which have been committed by the offender.¹⁰¹ The main difference between preventive sentences on the one hand and indefinite and additional sentences on the other is that preventive sentences relate solely to anticipated future harm, rather than, at least partly, to conduct which has already occurred.¹⁰² It is for this reason that only protective sentences are properly the inverse of suspended sentences. Accordingly, I will focus only on them.

In Australia there have been two separate pieces of legislation which have provided for protective sentences: the Community Protection Act 1990 (Vic)¹⁰³ and the Community Protection Act 1994 (NSW).¹⁰⁴ Both were ad hominem in nature: each was directed at a particular ‘dangerous’ individual.¹⁰⁵ The Victorian Act targeted Garry Webb (also known as Garry David) and the New South Wales legislation applied only to Gregory Kable.

The Victorian Act allowed for the preventive detention of Garry David for up to 12 months if the Supreme Court was satisfied that he presented a risk to the safety of any member of the public and that it was likely that he would commit any act of personal violence to any other person.¹⁰⁶ The NSW legislation provided that a court could order the preventive detention of Gregory Kable for up to six months where it was satisfied on the balance of probabilities that Kable was more likely than not to commit a serious act of violence.¹⁰⁷ This legislation was enacted in response to concerns that Kable, who was due for release after serving a sentence for the manslaughter of his wife, would harm relatives of the deceased whom he had sent threatening letters from jail. Multiple applications could be made for the detention of Kable; thus effectively he could be detained indefinitely.

The Victorian Act was repealed in 1993 following the suicide death of Garry David.¹⁰⁸ In Kable v DPP (NSW),¹⁰⁹ the New South Wales Act was ruled invalid by the High Court. By a four to two majority¹¹⁰ it was held that the Act violated the separation of powers doctrine¹¹¹ embodied in Chapter III of the Commonwealth Constitution because it conferred a non-judicial function on the Supreme Court by requiring the Court to participate in a process which was “far removed from the judicial process that is ordinarily invoked when a court is asked to imprison a person”,¹¹² and was so repugnant that it exceeded the outer limits of judicial power.

Each member of the majority had different reasons for striking down the Act. However, there were several features of the Act which the Court found particularly offensive. For one, it removed the ordinary protections inherent in the judicial process by permitting the deprivation of liberty without a finding of guilt for an offence,¹¹³ and enabling an opinion to be formed on the basis of material that may not be admissible in legal proceedings.¹¹⁴ Also, the outcome of any application appeared to be pre-determined by the Legislature, since it clearly was not envisaged that an order to detain Kable would be refused, and thereby the Act seemed to make the court an instrument of the Legislature.¹¹⁵ Finally, there was the ad hominem nature of the legislation.¹¹⁶

Thus the legislation in Kable was struck down by the High Court due to unique features of the Act which the Court believed infringed the separation of powers doctrine embodied in the Constitution.¹¹⁷ However, in my view, there is a more general objection to protective sentences, which stems from their incompatibility with the principle of proportionality. Before discussing the application of the principle of proportionality to protective (and suspended) sentences, I shall first consider the arguments which are normally levelled against protective sentences.

The most common objection to protective sentences relates to the notion of punishing people for crimes that have not been committed. This line of reasoning has been developed in several ways. It has been claimed that it is simply inherently unfair to punish in such circumstances.¹¹⁸ And, it has been argued that protective sentences are intuitively antagonistic to the notion of punishment: “one may promise punishment (or reward) for a future action, but to award it in advance would somehow seem to make it something else; a deterrent or incentive”.¹¹⁹ Despite the common sense appeal of such arguments, closer scrutiny reveals that their persuasiveness depends largely on which theory of punishment one adopts.

The first point to note is that a sound argument can be mounted that protective sentences are not intrinsically wrong, and that any intuitive unease towards them stems not from their perceived unfairness but from an underlying acknowledgement that human conduct can never be accurately determined in advance. If human conduct could be accurately predicted the intuitive disquiet about preventive sentences would in many instances readily dissipate. For example, if a person who was aware of the tragic events in Port Arthur, Tasmania, on 28 April 1996 when Martin Bryant killed 35 people went back in time to a moment shortly before the incident and had the opportunity to impose a protective sentence upon Bryant, it is doubtful whether many informed people would raise the slightest protest at the decision to imprison Bryant. Predictions about human behaviour will of course never become so accurate that such tragedies could be precisely forecast. However, fantastic examples such as this are helpful since they sharpen and illuminate the real premises and assumptions underlying our sentiments and conclusions.¹²⁰
(i) Utilitarian and Retributive Approaches to Protective Sentences

The suggestion that the above example shows that protective sentences are not inherently wrong is perhaps somewhat premature, since it may depend on which theory of punishment is being invoked. There are two main contemporary theories of punishment: utilitarianism and retributivism.¹²¹ The utilitarian theory of punishment regards punishment in itself as bad because it causes unhappiness to the offender. Punishment is only justified because of the wider contingent benefits it produces, which it is felt on balance outweigh the bad consequences. The suffering it inflicts on the offender is outweighed by the good consequences it produces by discouraging both the offender from re-offending and potential offenders from committing crimes in the first place, and thereby leading to a reduction in the frequency in which socially desirable laws are violated. If there are alternative forms of punishment which produce the same good consequences we must chose the one which imposes the least unpleasantness on the offender. On this view, the main benefit of punishment is the reduction of crime by deterring the criminal and others in the community and rehabilitating the offender, thereby reinforcing the wrongness of criminal behaviour providing a tangible sanction. Thus the utilitarian theory of punishment is forward looking: the commission of a criminal act does not justify punishment, instead punishment is only warranted because some good can flow from this. Clearly, from an utilitarian perspective there is no absolute obstacle to protective sentences; they are justified where this will increase net happiness. Thus, if we could be certain that a person would in the future commit an act resulting in immense suffering then net happiness would be advanced by imprisoning the potential offender.¹²²

However, this conclusion does not follow as surely from a retributive theory of punishment. It has been noted that retributivists are committed to the position that preventive sentences are necessarily wrong:

once an offender has undergone his ‘just deserts’ sentence, he has ‘paid his debt to society’ and is fully entitled to be released. To subject him to a further period of imprisonment is to punish him not for past offences, but for possible (and only possible) future offences.¹²³

Morris claims that preventive sentences are wrong even if we could be certain that the offender will offend in the future. He contends that people should be punished for what they have done, not for what they will or might do. People should be treated as responsible moral agents who can choose whether or not to commit future crimes, rather than treating them as ‘unexploded bombs’.¹²⁴

Wood rejects the proposition that the retributivist is necessarily committed to the inherent wrongness of preventive sentences, on the basis that retributivism only offers a theory of punishment, not a complete account of circumstances in which people can be forcibly detained. Wood argues that while protective sentences are unjustified, civil detention of dangerous offenders may be permissible on retributivist grounds as this has nothing to do with questions of desert but rather with social protection.¹²⁵ However, the distinction between a protective sentence and civil detention appears illusory. Civil detention still amounts to deprivation of liberty against one’s will. This constitutes a form of punishment on the basis of the definition adopted earlier: an unpleasantness imposed on a person. A feathered bird with a bill that quacks is a duck irrespective of what one chooses to call it. Lewis makes the point somewhat more eloquently:

To be taken without consent from my home and friends, to lose my liberty, to undergo all those assaults on my personality which modern psychotherapy knows how to deliver ... to know that this process will never end until either my captors have succeeded or I have grown wise enough to cheat them with apparent success - who cares whether this is called punishment or not.¹²⁶

Thus, given that it cannot be tenably asserted that forcible detention is not punishment, it may seem that the retributivist may be committed to denouncing protective sentences per se. This conclusion is, admittedly, complicated by the plethora of different retributive theories that have been advanced.¹²⁷ Retributive theories of punishment are not clearly delineated and it is difficult to isolate a common thread running through them.¹²⁸ All retributive theories assert that offenders deserve to suffer, and that the institution of punishment should inflict the suffering they deserve. However, they provide vastly divergent accounts of why criminals deserve to suffer.¹²⁹ Despite this, there are three broad similarities shared by retributive theories.¹³⁰

The first is that only those who are blameworthy deserve punishment and that this is the sole justification for punishment. Thus, punishment is only justified, broadly speaking, in cases of deliberate wrongdoing.¹³¹ The second is that the punishment must be equivalent to the level of wrongdoing.¹³² Finally, punishing criminals is just in itself:¹³³ it does not turn on the likely achievement of consequentialist goals. Punishment is justified even when “we are practically certain that attempts [to attain consequentialist goals, such as deterrence and rehabilitation] will fail”.¹³⁴ Thus is it often said that retributive theories are backward looking, merely focusing on past events in order to determine whether punishment is justified, in contrast to utilitarianism which is concerned only with the likely future consequences of imposing punishment.

The first of these similarities may appear to constitute a decisive argument against protective sentences in a retributive system of punishment. However, this is not necessarily the case if the concept of criminality is expanded slightly. While dangerousness in itself does not amount to a criminal offence, it does imperil the security of the community and because of this threat it could be argued that punishment is warranted. Viewed in this light the dangerous person is not innocent: through his or her behaviour he or she has caused social evil.¹³⁵ Such an approach is supported by the fact that retributivists have no difficulty with punishing people who engage in other types of conduct where the harm consists solely of threatening to violate the security of others. In this way, exhibiting tendencies, by words or conduct, which are viewed as potentially likely to lead to aggressive or harmful behaviour towards others could be classified as criminal behaviour and would be akin to offences such as stalking,¹³⁶ threats to kill or inflict serious injury,¹³⁷ and conduct endangering life or persons.¹³⁸ Thus even the retributivist is not necessarily logically committed to denouncing protective sentences and may therefore be willing to punish people purely on account of their dangerousness.¹³⁹

It should be noted that the above analysis applies irrespective of whether or not the person has committed previous acts of violence. To the contrary, Gross contends that protective sentences are only justified for those who have already committed offences, because by doing so they have breached their supposed social contract with the rest of society which provides that once a person commits an offence they forfeit certain rights and society can deal with them as it sees fit.¹⁴⁰ Not only is the existence of such a contract highly dubious, but it is unclear why the focus of the inquiry should be solely on past conduct, when the aim of protective sentences is to prevent future harm and/or curtail existing community unease about the prospect of such harm. Previous conduct is only one of many factors that may lead to a diagnosis of dangerousness - if one was aware in advance of the events of the Port Arthur massacre, whether or not Bryant had prior convictions would be totally irrelevant to a decision regarding the appropriateness of a protective sentence.
B. Inability to Predict Dangerousness

Thus it would appear that whichever theory of punishment one adopts, there is no fundamental objection to punishing people for crimes that they have not committed. Despite this, in my view, protective sentences are unjustified. The real objection to protective sentences lies not in their premature character (this is merely a matter of timing) but in our inability to confidently predict future human conduct. Given the complexity and unpredictability of human nature it is impossible to forecast future behaviour with any degree of certainty. Future promises, undertakings and declared intentions are one guide, but are far from conclusive. People change for the worse, but for the better as well. Behaviour is not only contingent upon fundamental values and beliefs, but also on the circumstances in which we find ourselves.

Although past conduct may be regarded as a powerful indicator of future propensities,¹⁴¹ and arguably basic values and predispositions are pervasive,¹⁴² current empirical evidence reveals that there is no reliable method for predicting dangerousness. Parke and Mason have noted that:

There is a wealth of material on the assessment of risk and the prediction of dangerous behaviour. But despite these vast outpourings, there are no reliable actuarial and statistical devices as yet that can predict with any degree of certainty the likelihood of dangerous behaviour.¹⁴³

The empirical evidence which does exist reveals a tendency to greatly over exaggerate the probability of future dangerous behaviour.¹⁴⁴ Few serious offenders commit other serious offences¹⁴⁵ and studies have shown that in predicting dangerousness, psychiatrists are wrong about 70 per cent of the time.¹⁴⁶ In Kable, Gaudron J described the prediction of dangerousness as “the making of a guess - perhaps an educated guess, but nonetheless a guess”¹⁴⁷ and McHugh J stated that it is “a prediction which can at best be but an informed guess”.¹⁴⁸ Curiously, while the psychiatric profession has repeatedly stressed the unreliability of psychiatric predictions of dangerous behaviour, the courts appear to be increasingly relying on them.¹⁴⁹

Thus it is impossible to be confident that a court which undertakes an inquiry into the dangerousness of an individual, using the best possible resources available, is likely to come to the correct decision.

Given this, the unease towards protective sentences stems not necessarily from the conviction that people should not be punished for crimes that they have not committed, but from the fact that we cannot predict with any degree of confidence that left to their own devices they would in fact commit serious offences in the future. This objection to protective sentences has a strong foundation in sentencing law: the principle of proportionality.
VI. The Principle of Proportionality

A. Statement of the Principle

In short, the principle of proportionality is that the punishment should fit the crime. It operates to restrain excessive, arbitrary and capricious punishment by requiring that punishment must not exceed the gravity of the offence,¹⁵⁰ even where it seems certain that the offender will immediately re-offend.¹⁵¹ Proportionality is not a justification for punishment, but rather a restraint on it.¹⁵² It is a principle which generally strikes a strong intuitive cord, and probably for this reason is found not only in sentencing law, but transcends many other areas of the law as well. As Fox notes, the notion that the response must be commensurate to the harm caused or sought to be prevented is at the core of the criminal defences of self-defence and provocation. It is also at the foundation of civil law damages for injury or death, which aim to compensate for the actual loss suffered, and equitable remedies, which are proportional to the detriment sought to be avoided.¹⁵³

Proportionality is one of the main objectives of sentencing¹⁵⁴ and the High Court decisions in Veen v R [No 1]¹⁵⁵ and Veen v R [No 2]¹⁵⁶ even went as far as declaring it to be the predominant objective of sentencing in Australia. In other jurisdictions it is treated just as importantly. For example, in relation to the Canadian sentencing system it has been noted that “the paramount principle governing the determination of a sentence is that the sentence be proportionate to the gravity of the offence and the degree of responsibility of the offender for the offence”.¹⁵⁷ In a similar vein, the White Paper forming the basis of the Criminal Justice Act 1991 (UK) declared that the aim of the reforms was to introduce a “legislative framework for sentencing, based on the seriousness of the offence and just deserts”.¹⁵⁸ Ultimately the Act did not expressly adopt these goals,¹⁵⁹ however the message was received in relation to the lengths of custodial sentences, the obligations of community sentences and the quantum of fines.¹⁶⁰