AustLII Home | Databases | WorldLII | Search | Feedback

University of Queensland Law Research Series

TC Beirne School of Law, The University of Queensland
You are here:  AustLII >> Databases >> University of Queensland Law Research Series >> 2007 >> [2007] UQLRS 1

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

Ratnapala, Suri --- "Reason and Reach of the Objection to Ex Post Facto Law" [2007] UQLRS 1; (2007) 1 (1) The Indian Journal of Constitutional Law, 140-168

Last Updated: 8 January 2009

THE UNIVERSITY OF QUEENSLAND
LEGAL RESEARCH SERIES


This article was originally published in The Indian Journal of Constitutional Law.


Suri Ratnapala, ‘Reason and Reach of the Objection to Ex Post Facto Law’, (2007) 1:1 The Indian Journal of Constitutional Law, 140-168


REASON AND REACH OF THE OBJECTION TO EX POST FACTO LAW
Suri Ratnapala[*]


ABSTRACT
This essay investigates the theory and scope of the constitutional principle against retrospective imposition of detriment expressed in the maxim nullum crimen, nulla poena sine lege. The theoretical bases of the objection are discussed as groundwork for an examination of key conceptual problems raised in the implementation of the principle. The logic of the argument for excluding civil deprivations from the scope of the objection is questioned and it is proposed that the civil-criminal distinction loses relevance in relation to past acts. The inadequacy of the ‘punishment’ test and the ‘legislative trial’ test are explained and the essay searches for a rational explanation for condemning some retrospective prophylactic laws while accepting others. The essay argues that the rule of law requires more stringent constraints on retrospective law making than are demanded by international instruments such as International Covenant of Civil and Political Rights and the European Convention on Human Rights and pleads that where positive constitutional law fails in this regard, nations should adopt them as constitutional practice.


KEYWORDS
Constitutional theory – rule of law - Ex post facto law – Recent US and Australian cases – kinds of objectionable retrospective laws


1. INTRODUCTION
On the 1st of October 2004, the High Court of Australia made two decisions that permitted the continued detention of persons already serving prison sentences for serious crimes. In the first decision, the Court rejected a challenge to Queensland legislation that authorised the State Supreme Court to order continued detention of a prisoner when it has reasonable grounds to believe that the prisoner poses a serious danger to the community.[1] In the second case, the High Court sanctioned New South Wales legislation that drastically limited the parol prospects of certain categories of offenders serving life imprisonment.[2] The two decisions expose the tenuous nature of the Australian public’s protection against ex post facto law. A troubling aspect of these decisions was the scant attention paid by the majorities to the ex post facto nature of these laws. Fardon and Baker were decided on the narrow issue whether the powers given to the State courts were compatible with their exercise of Commonwealth judicial power. The cases highlight the need for a re-examination of the reason and reach of the objection to ex post facto law and this essay is a contribution to that end.
It does not take much science to discover that the rule of law is impossible without strong restraints on legislative infliction of harm for lawful acts. Such impositions are not just assaults on the institutional structures that promote the rule of law but are direct negations of the rule of law. A law that undermines the separation of powers or representative government will expose the rule of law to subversion but will not per se defeat it. On the contrary, an ex post facto law in its classic form is the epitome of the capricious exercise of authority. The objection to ex post facto law is not founded on constitutional pragmatics but on the most fundamental demand of the rule of law, that the citizen is subject only to established and known law. Article 15(1) of the United Nations Covenant on Civil and Political Rights (ICCPR) condemns retrospective criminalisation of innocent acts and retrospective enhancement of punishment as does Article 7(1) of the European Convention on Human Rights (ECHR). The United Kingdom adopted the ECHR provision by the Human Rights Act 1998 and many other constitutional democracies have enacted the ban in this form.[3] This type of prohibition may be defeated by laws that impose pain in the guise of civil liability. I argue that the rule of law requires a more general constraint on retrospective imposition of harm than is prescribed by the ICCPR and ECHR standard and that where positive constitutional law fails in this regard, it should be recognised as a matter of constitutional principle and of political morality.
The term ex post facto law literally means ‘[arising] from past facts’ but in its technical sense refers to a class of laws that retrospectively inflict harm on persons on account of past legal conduct. The best known case of the ex post fact law is the bill of attainder and its variant the bill of pains and penalties by which the legislature directly imposes punishment on named individuals. However the class of laws that has been judicially condemned, particularly in the United States, and is found objectionable for the reasons to be discussed presently is wider. It includes laws that do not name individuals for retrospective harm but leaves their selection to other agencies. The term ex post facto law refers to this wider class.
The question whether in a given jurisdiction, limits on ex post facto law are requirements of positive constitutional law, constitutional convention or simply of political morality is important. My chosen aim though is to ask the questions what kinds of retrospective laws are objectionable and why they are objectionable. The answers will help lawyers determine what doctrinal means are available in a given constitutional system to restrain the enactment of such laws. If there is a one branch of law where the question ‘what is the law’ It must be emphasised that this is not a statement of the positive law of one or more countries although the United States and Australian law on the subject will receive close attention. It is not possible exhaustively to define the objectionable class without knowing all the novel devices that may be generated by legislators in the future. However, a non-exhaustive list of objectionable types will emerge from this discussion.
In Part 2 of this essay I consider the principal reasons for condemning ex post facto law. The discussion is mainly theoretical with judicial reasoning serving where relevant to explain the reasons. In Part 3, I engage some of the key conceptual and practical problems in identifying laws that attract condemnation for the reasons discussed. There is a heavy focus on landmark cases in the United States and Australia as a means of illustrating problems and where relevant judicial error. In Part 4, I itemize non-exhaustively the kinds of laws that are objectionable, if not in positive constitutional law, then in constitutional theory.


2. THEORETICAL OBJECTIONS TO EX POST FACTO LAW
Constitutional theory offers four reasons for the objection to ex post facto law. They overlap in some respects.


(1) Ex post facto law is not law
The idea of retrospective illegality makes any sense only if legality is understood as a provisional condition dependent on the future will of a legislative authority. If this is the case, the law’s primary function of providing guidance to conduct is severely weakened. Individuals who cannot predict the legal consequences of their actions cannot coordinate their behaviour in relation to each other. Therefore substantial restraint on the imposition of harm for past innocent acts is a necessary condition of social life.
The idea of a law conjures a general rule; hence the frequent question whether an enactment directed at punishing a single person or condemning a single transaction can properly be called a law. Blackstone wrote that an ‘act to confiscate the goods of Titius or to attaint him of high treason does not enter into the idea of a municipal law: for the operation of this act is spent upon Titius only and has no relation to the community in general: it is rather a sentence than a law’.[4] This rationale has received wide judicial recognition, the locus classicus being the US Supreme Court’s Opinion in Hurtado v California.

Law is something more than mere will exerted as an act of power. It must not be a special rule for a particular person or a particular case, but ... the general law ... so that every citizen shall hold his life, liberty, property and immunities under the protection of the general rules which govern society ... Arbitrary power, enforcing its edicts to the injury of the persons and property of its subjects, is not law, whether manifested as the decree of a personal monarch or an impersonal multitude.[5]

The Privy Council in Q v Liyanage invalidated an Act of the Ceylon Parliament calculated to lessen the prosecutorial burden and to enhance punishment in a particular treason trial on the basis that it was a legislative usurpation of judicial power.[6] The New South Wales Court of Appeal found legislation designed to defeat a trade union’s action challenging its deregistration to be an exercise of judicial, not legislative power. The court controversially upheld the legislation on the ground that the NSW legislature’s plenary power included judicial power.[7] This reason is untenable as the NSW legislature could not inherit by the language of the Constitution Act 1955 the kind of sovereign power that historically accrued to the UK Parliament. More recently, Justice Deane in the High Court of Australia observed that a bill of attainder ‘prohibits nothing, prescribes no rule of conduct and is incapable of being contravened since, by its terms, it is inapplicable to acts committed after its enactment’.[8]


(2) Nullum crimen, nulla poena sine lege
The maxim that there is no crime without a breach of the law is first a proposition of logic. If a crime is understood to be an act that is prohibited by law at the pain of a penalty, an act which is not so prohibited can never be a crime. The fact that Parliament visits the act with its retribution cannot make it a crime. The act is simply an event with reference to which Parliament elects to inflict pain on a person or group. However, the maxim nullum crimen sine lege is more than a proposition of logic. It is a substantive moral claim to humane treatment. Blackstone in his Commentaries observing that a person has no cause and cannot foresee a cause to abstain from doing what is legal wrote that in such cases ‘all punishment for not abstaining must be of consequence be cruel and unjust’.[9] The maxim nulum crimen sine lege also condemns law that eliminates an exculpatory defence that was available at the time of the commission of the act.[10] A person commits no crime if he has a legal excuse and the retrospective removal of the excuse amounts to the punishment of a lawful act.
The principle nulla poena sine lege which condemns the retrospective increase in penalties is again a substantive claim of justice. It can also be seen as a logical extension of nullum crimen sine lege. A retrospective increase in punishment is an infliction of new pain not prescribed by law. It is not punishment of the offence as the punishment has already been suffered or is being suffered. If there is a cause, it is the fact that the offender is a person who has been punished or is being punished according to law. Suffering lawful punishment is not itself unlawful.


Subtle ex post facto violations
In Collins v Youngblood, the Supreme Court declared that ‘subtle ex post facto violations are no more permissible than overt ones’.[11] This explains the Court’s extension of the nullum crimen, nulla poena sine lege principles to laws that ease the prosecutorial burden by retrospective changes to evidentiary rules. In Calder v Bull Justice Chase included within the class of impermissible ex post facto laws, ‘Every law that alters the legal rules of evidence, and receives less or different testimony than the law required at the time of the commission of the offence in order to convict the offender’.[12] Recently, in Carmel v Texas the Court ruled unconstitutional, a statute that retrospectively authorised conviction for certain sexual offences without corroboration of the victim’s evidence. The Court held that ‘a law reducing the quantum of evidence required to convict an offender is as grossly unfair as, say, eliminating an element of the offense, increasing the punishment for an existing offense, or lowering the burden of proof’.[13] In Stogner v California the US Supreme Court overrode a strong dissent to annul a California statute that allowed the resurrection of time barred prosecutions for child sex offences. The complaint against Stogner concerned events that happened 22 years before the charges were laid. The majority saw the statute of limitations primarily as a safeguard against conviction on insufficient or unreliable evidence and thought that its removal exposed Stogner to conviction on a lesser quantum of evidence than was required by law before the statute was passed.[14]
The reasoning here is more complex and contestable. The legislature is not overtly criminalising an innocent act but is denying a class of alleged offenders certain evidentiary safeguards. If the safeguards in fact decrease the probability of curial error their removal must logically increase the probability of innocent persons being punished, the kind of harm that the maxim seeks to prevent. However, as the dissent in Stogner pointed out, the theory founded on evidentiary concerns is weakened by the fact that the ban on ex post facto law as judicially accepted does not condemn laws that eliminate the time bar with respect to future offences or laws that extend unexpired limitation periods.[15] Hence if it is to hold, the theory must be further refined. The majority viewed the revival of expired prosecutions as a retrospective aggravation of the crime as condemned in Calder v Bull but so is the extension of unexpired limitation periods. More persuasively, the majority enlisted the notion of ‘reliance interest’ in aid of the theory. On the expiration of the time limit, persons have a reliance interest in the form of an expectation of immunity and hence may not preserve evidence of innocence.[16] The dissent ridiculed the reasoning with the observation that it involves a presumption that ‘criminals keep calendars so they can mark the day to discard their records or to place a gloating phone call to the victim’.[17]
Although diary keeping criminals may be hard to find, criminals who do not re-offend and who redeem themselves might not be so rare. In such cases the resurrection of expired prosecutions seems harsh. The point remains though that the extension of prosecutable periods or resurrection of time extinguished prosecutions as well as the retrospective easing of evidentiary burdens are not easily explained by the maxim nullum crimen, nulla poena sine lege.


(3) Failure of due process
This objection relates to bills of attainder and of pains and penalties. Whereas Blackstone denied these the status of law, John Lilburne saw in them the vice of the legislative imposition of punishment without trial. ‘To say that a freeman of England by Law may be tried by a Bill of Attainder, is irrational and unjust; for such a proceeding is no tryal, but rather a sentence, and is no act of jurisdiction, but an act of the legislative power, but no sentence can be past against an offender, but by some fore-declared visible, known rule of law; of which the supposed offender, either actually had, or might have had knowledge'.[18] This objection was articulated by the US Supreme Court in Selective Service System v Minnesota Public Interest Research Group when it stated that a bill of attainder is ‘a law that legislatively determines guilt and inflicts punishment upon an identified individual without provision of the protections of a judicial trial’.[19]
The doctrinal formulation of this objection is not based on the institutional separation of powers, but on a distinction drawn between the type of decisions that do not require the curial method and the type that do. The first type consists of decisions generating rules of general application and the second type comprises decisions that apply a general rule to a particular case.[20] The judicial method is unnecessary in the former case, but is essential to the latter. Tribe states that the distinction is ‘between those processes of choice which have such wide public ramifications that adversely affected individuals need not participate personally, and those choice processes which so focus upon particular persons that their personal participation must be assured’.[21]
Tribe’s formulation of the objection does not fully explain its rationale. The objection holds even if the legislature grants the individuals targeted a comprehensive hearing before it enacts the ex post facto law. A legislative hearing does not overcome this objection for two reasons. Firstly, as Cooley observes, the legislature ‘is not properly constituted to try with coolness, caution, and impartiality, a criminal charge, especially in those cases in which the popular feeling is strongly excited - the very class of cases most likely to be prosecuted in this mode’.[22] Secondly, even if it assumed that a representative legislature is impartial, it cannot act in the judicial mode without the guidance of pre-existing law. This highlights the fact that bills of attainder and of pains and penalties lack the qualities of both law and of adjudication but are cases of straightforward infliction of pain on chosen individuals.


(4) Ex post facto law offends the separation of powers doctrine
This objection is based on the claim that the enactment of certain forms of ex post facto law involves the exercise of judicial power. In countries where legislative and judicial powers are constitutionally reposed in separate organs, the legislative infliction of punishment or detriment on specified persons is unconstitutional as a matter of positive law. In England, where the Crown in Parliament had undivided power, the practice of enacting bills of attainder and of pains and penalties was abandoned as a matter of constitutional principle. It is worth noting that these Acts were not enacted to punish lawful acts but to punish criminal acts more expeditiously, more assuredly and more severely. The precedents collected by Hatsell[23] suggest that for the attainder procedure to be invoked, the party must have committed some act which by nature is a crime under the existing law. The practice (not always observed) was to allow the accused to defend themselves with counsel and witnesses before both Houses.[24] Sir John Hawles justified the practice on the ground that ‘it is no injustice for the supreme power to punish a fact in a higher manner than by law established, if the fact in its nature is a crime, and the circumstances make it much more heinous than ordinarily such crimes are’.[25] Even so, Parliament in these proceedings was not bound by law with respect to the substance of the offence, the procedural and evidentiary rules or the punishment. The last bill of attainder was enacted to attaint the Earl of Kellie and others in 1746 and the last bill of pains and penalties was enacted against Queen Caroline in 1820.[26]
The framers of the US Constitution reacted to the post independence wave of ex post facto laws that victimised loyalists[27] by expressly forbidding ex post facto laws in its first article. However, in many cases, the Supreme Court has treated the ban as part of the Constitution's implementation of the doctrine concerning the separation of powers. In Fletcher v Peck, the Court concluded that bills of attainder were ‘legislative judgments and an exercise of judicial power’. [28]In United States v Brown, the Court stated that the attainder clause was not intended as a narrow, technical prohibition, ‘but rather as an implementation of the separation of powers, a general safeguard against legislative exercise of the judicial function’.[29]
The view that the ban on ex post facto law is a necessary implication of the separation of powers is shared by the Privy Council and the High Court of Australia. The Privy Council made its definitive pronouncement in Q v Liyanage, an appeal from the Ceylon Supreme Court.[30] The constitution of Ceylon (now Sri Lanka) did not expressly prohibit ex post facto legislation. The two laws examined by the Privy Council constituted a legislative scheme designed to facilitate the trial, conviction and enhanced punishment of certain persons accused of an attempt to overthrow the lawfully established government of Ceylon. The Privy Council struck down the laws as repugnant to the separation of powers ordained by the constitution and observed that ‘if such Acts as these were valid the judicial power could be wholly absorbed by the legislature’.[31]
The early thinking in Australia was that the Commonwealth Constitution’s division of powers did not prohibit ex post facto law. Thus in King v Kidman,[32] a law that made the offence of conspiracy to defraud the Commonwealth an indictable offence with retrospective effect was questioned not on the grounds of attainder or separation of powers but on the basis that it did not fall within any of the enumerated the subjects of Commonwealth legislative power.[33] The argument found favour only with Chief Justice Griffith who found that in general ex post facto law was impossible to classify under any of the placita of powers vested in Parliament. The true category to which such laws belong he stated was ‘control over the liberty of the subject’ or ‘reward and punishment of citizens who deserve well or ill of the State’.[34] Nevertheless, he did not regard the impugned law as unconstitutional as it merely put into statutory form, an existing common law offence.[35]The other judges rejected the argument as a matter of construction stating that that the power to legislate on a subject contained the power to make laws having retrospective effect.[36]
Kidman stood for 75 years until the Mason Court in Polyukhovich v Commonwealth entertained the challenge to the War Crimes Act 1988 based on the separation of powers. Polyukhovich, an immigrant, was charged with war crimes that he allegedly committed during the German occupation of the Ukraine. The alleged acts were not punishable under Australian law although they would have been had they been committed in Australia. The charges were made possible by amendments to the War Crimes Act enacted in 1988. Polyukhovich challenged the Act on grounds, inter alia, that the Act was an invalid usurpation of judicial power of the Commonwealth. Five of the six judges agreed that the separation of powers doctrine barred the enactment of laws having retrospective penal effect[37] and Justice Dawson assumed it for the purpose of argument.[38] However, four of these judges concluded that the Act did not fall within the prohibited class of statutes. The common theme in their judgements was that a law that retrospectively makes an act punishable as a crime does not offend the separation doctrine provided it is general and not directed at specified individuals. Justice Dawson explained that under such law the ‘court is still left to determine whether an individual is guilty of having engaged in the prohibited activity, albeit an activity which took place before the law created the offence ...’[39] Justice McHugh observed: ‘Under such a law, it is still the jury, not the legislature which determines ... whether the accused is guilty or innocent of the charge against him or her’.[40]
As I argue later, retrospectivity necessarily involves a degree of specification of the persons targeted. That aside, the judges were plainly wrong in denying that the law was aimed at specific individuals. The preamble to the Act made it clear that its provisions were directed at persons who committed serious war crimes in Europe during World War II and who have entered Australia. The Act applied only to Australian citizens or residents accused of specified war crimes committed during the Second World War in the European theatre. (ss. 5, 9 and 11) Persons within this class were few and readily identified. It is also evident that the judges misunderstood the scope of the prohibition. Toohey J, echoed the other justices in concluding that ‘there is not a scintilla of difference’ between the roles of the judge and jury in a trial under this Act and the roles of the judge and jury in a trial under an identical law operating prospectively.[41] That was correct but was not the point in issue. The question was not whether the court was assigned a non-judicial task but whether the legislature had performed a judicial (or non legislative) function in selecting an identifiable group for its sanctions. Let us consider a less emotive hypothetical case. Statute A statute makes the sale of a standard loaf of bread at a price in excess of $1 an offence carrying a penalty of $500. It is prospective in operation and the legislature has not passed judgment. Compare this to Statute B made in 2005 that penalises all persons who in 2004 sold a standard loaf for more than $ 1. Under each law, the courts have the function of determining whether an accused person in fact sold bread at the prohibited price. However, in enacting the second law legislature has made a determination of its own - that all persons who in 2003 sold bread above the stipulated price have committed crimes. There is no escape for these persons. They remain liable to the penalty on conviction as long as the law is in force. As Deane J wrote in his dissenting judgment, such a law ‘prohibits nothing, prescribes no rule of conduct and is incapable of being contravened since, by its terms, it is inapplicable to acts committed after its enactment’.[42] It is this kind of law that current American jurisprudence regards as legislative judgments.
The Court could have upheld the law on the basis of the Nuremberg exception to the constitutional ban, namely the punishment of acts mala in se, or to employ the words of the ICCPR, acts that are ‘criminal according to the general principles of law recognised by the community of nations’.[43] It is a reasonable presumption that Polyukhovich, had he committed the alleged acts would have known that they were crimes. Hence, the law did not violate the assurance that the ex post facto ban provides ‘that no future retribution of society can occur except by reference to rules presently known’.[44] In contrast, the retrospective price control law punishes persons who had no idea that their actions were wrongful at the time they committed them.
The objection based on the separation of powers may be restated as follows. When the legislature enacts an ex post facto law, it does not act in a judicial manner, though sometimes such enactments follow a parliamentary inquiry. It does not determine the matter according to pre-established law, which is the hallmark of the judicial function. The legislature in such cases passes political judgment on particular individuals and in that sense subjects them to an inquisition without due process. Hence when we say that the enactment of such laws involves the exercise of judicial power, what we mean is that the legislature is making in a non-judicial manner, a decision that ought to be made by a court in a judicial manner.


THE PROHIBITED CLASS OF LAWS
I have discussed so far the reasons in constitutional theory for condemning ex post facto law but have not considered, except in very general terms, the types of law that are condemned. The following discussion addresses some of the conceptual difficulties attending the delineation of this class.


The chimera of the criminal-civil distinction
The ban in the US Constitution applies to any ‘Bill of attainder or ex post facto law’.[45] The words ‘ex post facto law’, unless superfluous must refer to laws other than bills of attainder. The question whether the ban applies only to retrospective punishment for crimes or extends to similar impositions of ‘civil’ deprivations is one that continues to challenge judicial minds.[46] In Calder v Bull,[47] the first case in which the question arose, the majority of the US Supreme Court concluded that the ban was limited to the retrospective punishment of crimes. Justice Chase thought that if the ban on ex post facto law was intended to apply to civil cases, the Fifth Amendment's ban on uncompensated taking of private property for public use would have been unnecessary[48] and Justice Paterson argued that the ban on contract impairment indicated ‘that the framers of the constitution ... understood and used the words in their known and appropriate signification, as referring to crimes, pains and penalties, and no further’.[49] Justice Iredell was emphatic that ‘The policy, the reason and humanity of the prohibition’ did not extend to civil cases or ‘cases that merely affect the private property of citizens’. The Judge thought, erroneously in my view, that such an extension would trump some of the most necessary and important acts of legislation.[50] According to this approach, the critical question is whether the law inflicts punishment for a crime or merely imposes a ‘civil liability’. The Supreme Court has never formally departed from this test.[51] Presently, I argue from the theoretical standpoint that this test is wholly misconceived. But first, it is necessary to show how the Supreme Court itself erased this distinction by its later rulings.
In dealing with retrospective imposition of civic disabilities, the Supreme Court has uncoupled ‘punishment’ from ‘crime’ and made retrospective punishment the criterion of invalidity. In two cases following the American Civil War, Cummings v Missouri[52] and Ex parte Garland,[53] the Supreme Court struck down laws that excluded persons from specified professions unless they swore that they did not take part in the rebellion against the Union. In Cummings, the complainant was a priest and in Garland, a lawyer. Each claimed that he was prevented by the ethics of his profession from taking the oath that was the condition for remaining in their profession. There was no question of criminality in these cases and the deprivations were not in the in the forms usually associated with crime namely: monetary penalty, confiscation of property, imprisonment or death. In Cummings the Court stated that the ‘deprivation of any rights, civil or political, previously enjoyed, may be punishment’, and included in this category disqualification ‘from the pursuits of a lawful avocation, or from positions of trust, or from the privilege of appearing in the courts, or acting as an executor, administrator, or guardian’.[54]
The unconstitutionality of this category of laws was confirmed in the leading 20th century cases on the attainder clause. In United States v Lovett[55]the Court struck down a federal law that prohibited the payment of future salary to three named government employees on the ground that in the past they engaged in subversive activities. The Court found that the provision constituted a ‘permanent proscription from any opportunity to serve the Government’ and ‘a punishment of a most severe type’.[56]In United States v Brown[57] the Court invalidated a law that made it a crime for a member of the Communist Party to hold office in a labour union during membership or within five years of the termination of membership. In Fletcher v Peck[58] the court ruled unconstitutional a law that rescinded a land grant considered to have been tainted with corruption although the current owner was innocent of wrongdoing. In Burgess v Salmon the Court invalidated a law that applied a tobacco tax retrospectively and allowed the penalty for non-payment to be recovered by a civil suit. The Court stated that the ban on ex post facto law cannot be evaded by giving a civil form to an essentially a criminal penalty.[59] These decisions extend the ban to enactments that impose deprivations having little resemblance to punishments traditionally associated with crime.
I argued previously that the notion of a retrospectively created crime makes no sense for a given meaning of crime. The legal concept of crime is of relatively recent origin. Until recently the law did not differentiate wrong into tort and crime. Wrongs existed as Winfield notes in the state of ‘viscous admixture’.[60] The genesis of crime is traceable to the start of Crown prosecutions. The Crown always prosecuted wrongs against itself such as treason. The reason why the Crown intervened to prosecute wrongs committed by subjects on other subjects is less clear though as Benson suggests, it gained financially by the forfeitures that resulted from convictions.[61] Later as the state progressively took on the role of social and economic regulator and provider, parliament legislated to create crimes in furtherance of policy.[62]
Blackstone’s distinction between mala in se (acts wrong by their nature) and mala prohibita (acts prohibited by the state) remains useful. [63] Mala in se are acts that are wrongs by the moral values of the community. In England these were established as legal wrongs through the build up of common law precedent. The idea of a retrospective offence mala in se makes no sense at all. There can be technical reasons why a particular incident of mala in se escapes punishment. The most dramatic examples concern war crimes or crimes against humanity where defendants often advance the defence of lawful orders. Crime committed beyond the limits of national jurisdictions offer other examples. Retrospective legislation to provide for redress or punishment in these cases is not considered repugnant to the ban on ex post facto law. This is the ‘Nuremberg principle’ implemented by international treaty law on human rights. Thus Article 17.2 of the ICCPR and 7.2 of the ECHR permit ‘the trial and punishment of any person for any act or omission which, at the time when it was committed, was criminal according to the general principles of law recognised by civilised nations.’[64] The logic of this exception is apparent. Persons who commit these kinds of acts usually know that they are serious wrongs against the general norms of society. Hence retrospective trial and punishment of such acts do not defeat legitimate expectations but only the hopes held by wrongdoers of getting away with their heinous acts.
Mala prohibita raise by far the more difficult issues. In theory, the state can make any act a crime in the sense of mala prohibita. As Lord Atkin observed, ‘The domain of criminal jurisprudence can only be ascertained by examining what acts at any particular period are declared by the State to be crimes, and the only common nature they will be found to possess is that they are prohibited by the State and that those who commit them are punished.’[65] There is always the problem of circularity in defining a crime. Glanville Williams regards a crime as a ‘legal wrong that can be followed by criminal proceedings which may result in punishment’.[66] A crime by this definition is an act declared by the legislator to be a crime entailing punishment. Williams denies that this is a circular definition pointing out that a criminal proceeding is one that attracts special procedural and evidentiary rules not applied in civil proceedings.[67] Williams is mistaken to think that the attachment of procedural and evidentiary requirements cures the circularity of the definition. It simply means that the law maker can make any act a crime which then attracts the procedural and evidentiary constraints. The central element of Kenny’s definition of a crime is harm caused by human conduct that the sovereign power desires to prevent by means including the threat of punishment. He also notes that the prosecution of crimes attract legal proceedings of a special kind.[68] All these definitions postulate that a crime in the modern sense is any act designated as a crime by the state but by the same token they indicate that the state cannot convert past innocent acts into crimes, as crime must concern a prohibited act. A retrospective crime makes sense only if crime is differently defined as any act, past or future, that the state chooses to punish whether or not it is lawful. Such a definition carries the monstrous proposition that the state may at its discretion lawfully inflict pain on any person for doing what is lawful.
If it makes little sense to talk of retrospective crimes, it makes even less sense to speak of retrospective civil liability. The most important feature of civil liability is that it is not punishment but reparation. Civil wrongs give rise to obligations to repay debts, to compensate for loss caused by wilful or negligent actions or breach of contract, to render specific performance of contractual undertakings or to effect restitution of unjust gains. In each case the obligation is to make reparation, to restore the party harmed, as far as possible, to the position before the wrong was committed. A person cannot incur such an obligation if he has acted lawfully. Where there is no breach of obligation, there is no question of civil liability and the legislative imposition of detriment amounts to the infliction of pain on innocent citizens. Hence, the civil-criminal distinction is unsustainable as a test of unconstitutional attainder. Lehmann proposes that the proper question is whether the statute is punitive or regulatory in nature,[69] but this test has its own serious limitations.


Punishment – a useful but insufficient test
The orthodox view is that the ban prevents ex post facto imposition of ‘punishment’ but not regulatory devices.[70] Punishment in its ordinary sense is not synonymous with harm but is associated with response to wrong doing. Punishment may be motivated by retribution, rehabilitation, prevention, deterrence or a combination of them.[71] In the case of laws that offend nulla poena sine lege the increased punishment relates to previous wrongdoing. Strictly speaking laws that offend nullum crimen sine lege do not impose punishment but inflict of harm. Infliction of pain for an innocent act is punishment only in a perverse sense. A child is punished for behaving badly and a criminal is punished for committing a crime. On the contrary, a sadist who inflicts pain on a victim or a robber who takes property does not impose punishment but causes wrongful harm. Likewise a legislature that imposes detriments on selected individuals out of spite or for political gain or indeed in the prosecution of state policy does not impose punishment but causes intentional harm to a selected individual or group. It is evident that punishment in the conventional sense is an inadequate concept on which to enforce the principle nullum crimen sine lege. The more general concept of detriment better explains the operation of this rule.
The types of detriment that has attracted the ban include disqualification from office,[72]forfeiture of property,[73] disenfranchisement,[74]and retrospective tax penalty.[75] As against these, the Supreme Court has upheld laws that take away the right of convicted felons to practice medicine[76]and to be employed in waterfront labour organisations[77]although the laws clearly operated to impose detriments in addition to penalties already prescribed by law. In an influential student comment in the Yale Law Journal[78]Dershowitz urged the abandonment of the ‘punishment’ test saying that it involves inexact and emotive distinctions that in some cases are difficult to achieve. He proposed instead that the attainder clause should be understood as banning legislative trial and not legislative punishment.[79] Dershowitz asks how the following two statutes can be distinguished. (1) ‘No person afflicted with a contagious disease shall teach school’ and (2) ‘John Jones, because he has a contagious disease, shall not teach school’. Only the second law is contrary to the attainder ban, but Dershowitz argues it is no more punitive than the first law. He says that the second statute offends the attainder clause not because of the legislature’s intent to punish but because of the legislature’s application of its general legislative mandate to a specific individual. Dershowitz’s assessment that the two laws are equally detrimental is questionable. Firstly, the under the second law Jones cannot avoid the detriment by showing that he is free of disease and secondly the first law does not discriminate against him whereas the second does. The test of legislative trial remains inadequate even when the two statutes are assumed to be equally detrimental.


Legislative trial - an inconclusive test
The first and obvious point to make is that when the legislature selects a person or group for punishment, it does not always conduct a trial. The most abhorrent instances of ex post facto law are not those where a person is accused and tried by the legislature but those where the legislature punishes without trial, where no charges are laid and no defence is heard. It is unhelpful to equate the enactment of these laws to trials as they represent the classic cases of punishment without trial. A law that retrospectively punishes all persons who supported the Communist Party will fail the attainder clause not because of legislative trial but because the target group are identifiable and are denied a legal way of avoiding punishment. The legislature in this case has not ‘tried’ this group but has determined as a matter of policy that they ought to be punished if indeed the courts find them to have supported the Communist Party. In United States v Brown the Supreme Court stated that ‘Congress may weed out dangerous persons from the labour movement but must do so by rules of general applicability. Congress possesses full legislative authority, but the task of adjudication must be left to other tribunals’.[80] A necessary implication of this injunction is that the legislature also cannot impose detriment as an arbitrary projection of its power.


The problems with the ‘punishment versus regulation’ test
The US Supreme Court has sought to exempt from the attainder clause, regulatory or prophylactic measures that have no punitive end but retroactively defeat vested rights. This test like the others discussed, fails to account for all of the Court’s decisions. The Supreme Court has upheld laws that took away the right of convicted felons to practice medicine and to be employed in waterfront labour organisations in addition to penalties already prescribed by law. These laws were regarded as prophylactic although permanent exclusion from occupations is a known form of punishment. Conversely, in United States v Brown, a purportedly prophylactic measure involving the exclusion of past communists from labour organisations was struck down on the ground that the legislature had thereby determined that past membership in the Communist Party made persons unsuitable to engage in designated occupations.
How is the exclusion of communists from public office different to the exclusion of persons with contagious diseases from school teaching or grand mal epileptics from driving? As Dershowitz points out, in the latter cases the legislature does not have to engage in a ‘trial’ since the danger to society is conveyed by the established meaning of the term employed to describe the group.[81] A contagious disease is communicated on contact. Grand mal epilepsy results in sudden fits. The effects of these disabilities are so well known that the only question for a court is whether the excluded person suffers the disability.
The corollary of the last discussed proposition is: where the danger to society from the disability is not palpable, the exclusion without judicial trial of the issue amounts to unfair treatment and hence a form of detriment without trial. In the law which proscribes sufferers of grand mal epilepsy from driving motor vehicles, the words ‘sufferers of grand mal epilepsy’ is actually shorthand for the condition which makes persons incurably prone to unpredictable and uncontrollable fits. Behind the apparent specificity of the words lies a rule of such generality that its enactment properly belongs to the legislative branch. In contrast, as the Supreme Court states in Brown it is a fallacy ‘that membership in the Communist Party, or any other political organization, can be regarded as an alternative, but equivalent, expression for a list of undesirable characteristics’.[82] The law that disenfranchises communists does not enact a general rule but imposes an arbitrary sentence.
The effect of United States v Brown is that a law will be treated as regulative only if: (a) it is not retributive in aim or motive and is not punishment in the conventional sense (b) does not arbitrarily select persons for detriment and (c) does not adjudge an individual or group to be dangerous to society without a judicial except when they suffer a condition that is the semantic equivalent of the danger.[83] Many laws previously upheld by the Supreme Court would have failed this test. In particular, the test throws in serious doubt, the cases concerning the disqualification of convicted felons.[84] It is arguable that ‘convicted felon’ is not the semantic equivalent of a person with undesirable character, as it is not universally accepted that a person who commits one felony is incapable or redemption and reform.
If this approach is accepted, it calls into question the practice of legislative impositions based on findings of a tribunal that does not follow the curial process. In these cases the legislature does not conduct a trial but imposes detriment without granting the affected persons the benefit of a judicial trial. In the Ceylon case of Kariapper v Wijesinha,[85]the Privy Council considered whether a law that imposed civic disabilities (including expulsion from Parliament and disenfranchisement) on the basis of findings of corruption reported by a Royal Commission of Inquiry, violated the attainder ban implied in that country's constitution. The judges rejected the challenge on two grounds. They held that the imposition of the civic disabilities was not punishment but a measure to ‘keep public life clean for the public good’.[86]On the question of legislative judgement, their Lordships stated that ‘it is the commission's finding that attracts the operation of the Act’ and that Parliament ‘did not make any findings of its own’.[87] This decision is questionable as Parliament in adopting the commission's findings arguably made a judgment of its own. The commission did not follow the normal rules of evidence and criminal procedure and did not convict the plaintiff of any criminal offence. His civic rights were extinguished by a political act following an extraordinary process established ex post facto.


Can the ban on attainder apply to prospective provisions?
An ex post facto law visits persons with detriment on account of past conduct. Dershowitz in the Yale Law Journal comment argued that that retrospectivity is not essential for a law to be called a bill of attainder. He points to the example of the Act for the Attainder of the pretended Prince of Wales of High Treason 1700 that besides attainting the prince made it treason for a person to correspond with him in the future.[88] The example in my view is unsound. The Act was prospective insofar it created a new obligation on subjects to refrain from corresponding with the prince. However, it was retrospective in relation to the prince himself. It imposed on him an additional retrospective detriment on account of past conduct namely, the loss of his freedom of communication by correspondence. As I argue presently, retrospectivity occurs whenever an individual is selected for punitive treatment.
Dershowitz’s concern was that the requirement of retrospectivity may defeat the object of the attainder ban. It was fuelled by certain decisions of the Supreme Court that validated penalties imposed on specified persons on the ground that the penalties serve to prevent future harm that may be caused by such persons. However, these decisions do not stand up to close scrutiny and since Dershowitz wrote his essay, the Supreme Court has disapproved of them. The main offending case is American Communications Association v Douds,[89]in which the Supreme Court considered the validity of a law that disadvantaged labour organisations whose officers had not filed the so called ‘non-communist’ affidavits. This precedent was overturned by the Supreme Court in United States v Brown.[90] It is hard to find a clearer example of Congressional judgment of guilt on account of past conduct, than the law considered in Douds. Congress having investigated the activities of communists determined that they were not fit to hold office in labour unions. The disqualification from office was based on past conduct. This is the view that the Supreme Court reached in Brown when it found that the law disqualifying communists was ‘to purge the governing boards of labour unions of those whom Congress regards as guilty of subversive acts and associations and therefore unfit to fill the positions which might affect interstate commerce’.[91]


Specification causes retrospective effect
Specification of persons for detriment may be done by naming the persons or by defining a class in a way that enables the identification of the persons targeted for detriment. A law that directly inflicts detriment on specified persons is necessarily retrospective in effect. How so? The critical distinction to notice here is between prescription of detriment and infliction of detriment. While detriment can be prescribed for future conduct, it is impossible to inflict detriment on account of future conduct that may or may not occur. If detriment is inflicted on specified persons, it takes effect irrespective of what happens in the future. Therefore, the reason for inflicting the detriment must relate to the past. The motivator may be past reprehensible conduct or there may not be a discernible reason for inflicting pain. If it is the latter, the pain is inflicted on the persons for being who they are and who they have been. Either way, the law is retrospective in operation.
What if the law that selects specified persons is prophylactic? Let us consider the following laws. Law 1 states that ‘any person who in the opinion of the mental health tribunal is likely to engage in harmful behaviour may be interned by order of court’. Law 2 states that ‘If in the opinion of the mental health tribunal Titius is likely to engage in harmful behaviour he may be interned by order of court’. Both laws promote public safety. Law 1 is prospective and defensible against the attainder ban. Law 2 is not as it makes Titius uniquely liable to its process leaving out others in that class. Assuming that this is detriment, there is no reason for inflicting it except that Titius is who he is and has been.


Judicial detention orders - the dangerous prisoner cases
Deprivation of liberty by detention is a central case of punishment and usually involves the aims of retribution or deterrence. Hence the detention of prisoners beyond their initial sentence presumptively offends the principle nullum crimen, nulla poena sine lege. There are certain types of detention that are considered outside the ban. They include detention on grounds of mental illness, communicable disease, illegal immigration, national security, criminal investigation and remand pending trial. Security related detention is usually executively determined while detention on health grounds and remand are judicial acts. Except in the problematic case of national security detention, detainees are treated differently from prisoners on punishment and are kept subject to strict conditions concerning treatment and release.
In Kansas v Hendricks, the US Supreme Court unanimously found the detention provisions of the Kansas Sexually Violent Predator Act did not violate the substantive due process, ex post facto and double jeopardy provisions of the Constitution. The Act was saved by its limited scope and its panoply of safeguards. The Act authorised ‘civil commitment’ of convicts who, due to ‘mental abnormality’ or ‘personality disorder’ are found likely to engage in ‘predatory acts of sexual violence’. In the Opinion delivered by Justice Thomas the Court ruled that ‘a finding of dangerousness, standing alone, is ordinarily not a sufficient ground upon which to justify indefinite commitment’ and observed that civil commitment has been allowed when statutes have combined ‘dangerousness’ with an additional factor such as ‘mental illness or mental abnormality’.[92] The Kansas statute required proof of the mental condition beyond a reasonable doubt, limited detention to one year at a time and provided effective recourse against orders.
In sharp contrast, the Dangerous Prisoners (Sexual Offenders) Act 2003 (Qld) considered in Fardon v Attorney-General (Qld) dispenses with the mental health qualification and allows the court to impose continuing detention on a prisoner if it is satisfied that there is serious danger in the form of ‘an unacceptable risk that the prisoner [would] commit a serious sexual offence’.(s.13(2)) The court may take account of the prisoner’s criminal history and while it may consider medical, psychiatric and psychological reports it need not come to a finding of mental illness.(s.45(4)) The detainee is deemed to remain as a prisoner (s.14(2)) and as the dissenting Justice Kirby stated, ‘After the judicial sentence has concluded, the normal incidents of punishment continue’.[93] The majority did not disagree with this characterisation but focused on the prophylactic aspect of the detention. There is no reason to doubt the protective goal of the law. Yet, a protective measure can also be punitive, especially when it is disproportionate or ill adapted.[94] Such a law offends the ex post facto principle in two ways. It exposes future offenders to multiple punishments for the same offence and it makes serving prisoners retrospectively liable to additional punishments.
Whereas the Queensland law allowed extended imprisonment by further detention, the Sentencing Act 1989 (NSW) considered in Baker provided for the denial of parol rights to prisoners of a defined class who are subject to ‘non release recommendations’ (euphemism for never to be released) by the trial judge. These prisoners are not eligible to parol unless the State Supreme Court finds ‘special reasons exist that justify’ the making of parol orders. Much time was spent on the question whether ‘special reasons’ imported an intelligible standard. The majority found that it did but their laboured case has no bearing on the question whether this law offends the principle nulla poena sine lege.
Neither statute was challenged on grounds of ex post facto effect. The reason was the High Court’s long held view that State parliaments possessed indivisible plenary power within their jurisdictions including the competence to enact ad hominem law and ex post facto law. Counsel for the prisoners relied instead on the much narrower ground established by Kable v DPP.[95] The High Court in that case struck down the Community Protection Act 1994 (NSW) which was designed to secure the further detention of a single named prisoner for public safety reasons. Confronted by its own dogma concerning the plenary power of State parliaments, the Court found means of invalidating the statute in the implications of the separation of powers doctrine in the federal Constitution. Observing that State courts were integral parts of the federal judicial hierarchy the majority concluded that the power to detain a named individual was incompatible with the court’s exercise of federal judicial power and hence could undermine public confidence in the federal judicature in a way that offended the separation of powers in the federal Constitution.[96]
In Fardon and Baker the Court regarded the more general power to impose preventive detention on classes of prisoners as compatible with the federal judicial role of state courts. It is possible to take the contrary view. Under the legislation examined in Kable, the court’s power was exhausted with one prisoner whereas under the laws considered in Baker and Fardon the court has continuing authority to make detention orders with respect to a class. It is arguable that if the law in Kable was bad for the exercise of federal judicial power, the laws in Baker and Fardon are worse. Each of these laws would presumably fail if enacted as federal legislation enlisting federal courts. They would have failed for investing non-judicial powers in federal courts contrary to the rule in Boilermakers’ Case.[97]They would also fail on the nullum crimens, nulla poena sine lege rule under the authority of Polyukhovich. The High Court’s position regarding State courts as revealed by the judicial detention cases is that ex post facto State laws are not unconstitutional except to the extent that they compromise the federal judicial role of State courts.


Legislative intervention in judicial proceedings
Judicial proceedings are instituted to vindicate rights and enforce duties. Hence every law that alters the rights in issue before a court imposes detriment on a party with retrospective effect and hence prima facie invites condemnation for ex post facto effect. However, if the intervening law is purely prophylactic in aim it may not offend the ban. Take the case of the grand mal epileptic whose application for a driving licence is rejected. If the law does not permit the licensing authority to deny the sufferer a licence on this ground and its decision is challenged, the public interest may require the disqualification of grand mal epileptics generally and further validate the refusal of licences to such persons in the past. Such a law is purely prophylactic although it may impact on pending cases. It does not select persons in similar conditions for dissimilar treatment and has no punitive intent.
In Nelungaloo Pty Ltd v Commonwealth,[98]the High Court considered the validity of an order for the acquisition of wheat made under a war time regulation.[99]There were serious doubts on the question whether the law authorised the acquisition of future as opposed to existing crops. The validity of the acquisition order was challenged but while the case was pending Parliament enacted legislation to clarify the law and to validate orders already made.[100] The plaintiff argued, inter alia, that the retrospective validation of the regulation under challenge in the court was a usurpation of judicial power. It was critical to the decision that the acquisition order was not directed exclusively at the wheat grown by the plaintiff, but applied to the entire Australian harvest. The only judge to consider the question, Dixon J, dismissed the objection stating: ‘It is simply a retrospective validation of an administrative act and should be treated in the same way as if it said that the rights should be the same as they would be, if the order was valid’.[101] The law would have survived a challenge on ex post facto grounds as it was prophylactic in nature, had no punitive aim and did not involve the legislature in a ‘trial’.
In Nicholas v The Queen[102] the High Court upheld the Crimes Amendment (Controlled Operations) Act 1996 (Cth) that required courts to disregard offences committed by official agents in anti-narcotics operations in exercising discretion to admit evidence of importation. The enactment followed the High Court’s re-iteration, in Ridgeway v The Queen, of the public policy discretion to exclude evidence of official wrong doing.[103] Nicholas did not claim that the amendment would affect the jury’s verdict but argued that the law infringes or usurps the judicial power of the Commonwealth in two ways. Firstly, it was contended that the limitation of the public policy discretion denied courts the authority to protect their integrity and dignity which is an attribute of judicial power. Although the Justices McHugh and Kirby agreed, the majority viewed the limitation as one of procedure and evidence and hence ultimately within legislative power.[104] Secondly, it was argued for Nicholas that the law, though facially general, was directed at a small group of known persons who were subject to ‘controlled operations’, hence it was similar to the usurpation of judicial power condemned in Liyanage. The argument found no favour with the majority mainly due to the prospective operation of the law.
Whereas in Nelungaloo and Nicholas the legislation was general and impacted incidentally on a pending case, the law challenged in The Queen v Humby; Ex parte Rooney[105] was directed at specified judicial decrees. It had the aim of extinguishing the constitutional right of the plaintiffs to have their cases under Commonwealth law heard by a judge. Section 72 of the Australian Constitution allows Parliament to vest federal judicial power in State courts. In Knight v Knight,[106] the High Court had ruled that the constitutional separation of powers required that federal judicial power devolved on State courts must be exercised by judges of the State court and that the Master of the Supreme Court of South Australia was not a judge. One effect of the decision was to call in question all the maintenance orders made by the Master under section 84 (1) of the Matrimonial Causes Act 1959 (Cth). The Commonwealth Parliament responded by enacting the Matrimonial Causes Act 1971 with the sole purpose of validating the Masters' decrees notwithstanding Knight v Knight. The Act was promptly challenged. In my view, the Act was patently unconstitutional as it sought to validate specific orders that had been judicially determined to be unconstitutional. The judges who addressed this issue engaged in exercises of mind boggling casuistry to deny that the Act had this effect. McTiernan J claimed that the impugned decrees were validated not as judicial decrees but as legislative enactments.[107]Stephen J maintained that the orders of the Master remained ineffective but ‘the sub-section operates by attaching to them ... consequences which it declares them to have always had and it describes those consequences by reference to the consequences flowing from the making of decrees by a single judge of the Supreme Court of the relevant State’.[108]Mason J declared that the order of the Master does not acquire validity ‘merely because the statute attributes to it the effect it would have had, had it been a judicial determination’.[109] The judges evidently saw a distinction between validation of an order and the statutory attachment to an invalid order of consequences which the order would have generated had it been valid. The legal effect though was exactly the same. The critical differences between the laws considered in Nelungaloo and Humby is that the former was general and prophylactic whereas the latter was specific and non-prophylactic. In fairness it must be noted that the Matrimonial Causes Act 1971 had no sinister design and had no punitive intent. It was enacted to cure a defect that resulted from earlier understanding of the effect of s. 72 of the Constitution that federal jurisdiction vested in State courts could be exercised by officers of the court. Nevertheless the Act denied certain parties of a constitutional right albeit one they did not know they had until Knight v Knight.
In contrast the law considered in Australian Building Construction Employees’ and Builders Labourers’ Federation v Commonwealth (Cth BLF Case)[110] was straightforwardly punitive. The law was made for the sole purpose of destroying the status of the BLF as a registered trade union. The BLF was considered a rogue union by the Commonwealth and State governments and by the Australian Council of Trade Unions (ACTU). BLF was registered under the Conciliation and Arbitration Act 1904. Registration conferred an extensive range of rights and privileges. The Act provided for two methods of deregistration, one judicial, and the other quasi-judicial. The judicial mode led to a determination by the Federal Court on objective criteria set out in the Act.[111]The quasi-judicial method culminated in a determination of the Conciliation and Arbitration Commission based on a more broadly expressed set of social concerns.[112] In its determination to strip the BLF of its statutory status, the federal government secured the passage of the Building Industry Act 1985 that made special provision for deregistering the BLF. The Act empowered the Commission on the application of the Minister, to make a declaration that the BLF or an officer or employee thereof had engaged in industrial action. Following such a declaration, the Minister was authorised on public policy grounds to order the cancellation of registration and to impose certain other deprivations on the BLF. The BLF and its members were selected for harsher treatment under the law. The Act was would have been struck down in the US even on the most stringent construction of the attainder ban. Yet, it was challenged in the High Court only on the narrow ground that it was not within the ‘conciliation and arbitration’ power set out s. 51 (xxxv). The court rejected the argument.[113]
The Commission duly made the requisite declaration authorising the Minister to deregister the BLF. The BLF went back to the High Court complaining of a denial of natural justice and sought orders to quash the declaration and to prohibit the Minister from ordering the deregistration of the union. While this case was pending Parliament passed two more laws, the Builders Labourers' Federation (Cancellation of Registration) Act 1986 and the Builders Labourers' Federation (Cancellation of Registration - Consequential Provisions) Act 1986. The first Act directly cancelled the registration of the BLF and the second Act imposed consequential disabilities on the officers and members of the union. BLF returned to the High Court complaining that the two Acts amounted to an exercise of judicial power and ... an interference with [the] Court's exercise of the power ...'.[114] The challenge was unanimously rejected on the premise that the legislation simply deregistered the Federation, thereby making redundant the legal proceeding and observed: ‘It matters not that the motive or purpose of the Minister, the Government and the Parliament in enacting the statute was to circumvent the proceedings and forestall any decision which might be given in those proceedings’.[115] The ex post facto nature of the law escaped the court’s attention. In enacting these laws, Parliament itself adjudged that the BLF was deserving of punishment in the form of deregistration. It is hard to find a clearer case of a legislative trial and punishment in modern times.


4. SUMMARY
The discussion of the theory and practical reasons for objecting to ex post facto law exposes the following types of law to condemnation.

  1. Ad hominem laws by which the legislature directly imposes punishments on named individuals for existing offences with or without a trial by the legislature. The bills of attainder and of pains and penalties typify this class but it includes laws that impose detriments on specified individuals on the basis of adverse findings of tribunals where the tribunal is not required to determine unlawful conduct according to established law.
  2. Laws that retrospectively increase punishments for existing offences.
  3. Laws that do not directly punish persons but which create new liabilities for past conduct as judicially determined. Such liabilities may be criminal or civil in nature and will include laws that impose civic disabilities. This category does not include laws designed to bring to trial persons accused of committing acts mala in se or, to use the words of the ICCPR ‘criminal according to the general principles of law recognised by the community of nations’.
  4. Laws that retrospectively remove defences or exceptions to civil or criminal liability.
  5. Laws that in respect of a class of offences lessen the prosecutorial burden by retrospectively modifying procedural or evidentiary rules.
  6. Laws that impose future non-prophylactic obligations on selected individuals. The arbitrariness of the selection reveals punitive intent and for reasons explained, the selection is necessarily based on past events.
  7. Laws that are facially prospective and prophylactic but which select for its attention some but not all agents thought to be the source of potential harm. If named HIV positive persons are excluded from an occupation while others with the condition are not, the law will be open to the objection that it is imposing the disability on persons for who they are rather than for what they cause.
  8. Laws that select a class of persons for the imposition of disqualifications with reference to a named impairment when the name of the impairment is not the semantic equivalent of a universally recognised and relevant disability.
  9. Laws that affect the outcome of pending legal proceedings by dictating the decision or by retrospectively altering the rights of the parties. Such a law will escape condemnation if it applies generally or to all members of a class and has a purely prophylactic effect.

It is not surprising that ex post facto laws of the kind criticised in this essay are uncommon in countries that maintain acceptable levels of constitutional government. Yet, the attraction of this type of law as means to short term ends, both good and bad, is ever present. It is tempting to think that democracy is a sufficient safeguard against gross abuse of ex post facto law but history cautions against such faith. Hence even where constitutions place no formal limits, the rule of law demands that lawmakers adopt restraints on retrospective legislation as a matter of constitutional practice. Constitutional practice owes much to the pressure of public opinion and I hope that this discussion will be a helpful contribution to public understanding of the reason and reach of the objection to ex post facto law.



[*] Professor of Law and Director, Centre for Public, International and Comparative Law, T C Beirne School of Law, University of Queensland.
[1] Fardon v Attorney-General (Qld) [2004] HCA 46; (2004) 210 ALR 50.
[2] Baker v R [2004] HCA 45; (2004) 210 ALR 1.
[3] See for example New Zealand Bill of Rights Act 1990 s 26(1); Canadian Charter of Rights and Freedoms s 11(g); Constitution of India Art 20, German Grundgesetz Art 103.
[4] Sir William Blackstone, Commentaries on the Laws of England vol 1, (London: T. Cadell & W. Davies, 14th ed, 1803) 44-45.
[5] [1884] USSC 84; 110 US 516, 535-536 (1884).
[6] [1967] 1 AC 259
[7] Building Construction Employees’ and Builders Labourers’ Federation (NSW BLF Case) (1986) 7 NSWLR 372. See my criticism of this decision in S Ratnapala, Australian Constitutional Law: Foundations and Theory (Melbourne: Oxford University Press, 2000) 112-114.
[8] Polyukhovich v Commonwealth [1991] HCA 32; (1991) 172 CLR 501, 631.
[9] W Blackstone, n 2 above, 46.
[10]Kring v Missouri [1883] USSC 102; 107 US 221 (1883); Dobbert v Florida [1977] USSC 115; 432 US 282 (1977); Carmel v Texas [2000] USSC 37; 529 US 513 (2000).
[11] [1990] USSC 113; 497 US 37, 46 (1990)
[12] [1804] USSC 10; 3 Dall. 386, 390 (1798).
[13] [2000] USSC 37; 529 US 513, 532 (2000).
[14] [2003] USSC 4778; 539 US 607, 615.(2003)
[15] Ibid 650.
[16] Ibid 631.
[17] Ibid 670.
[18] Quoted from W.B. Gwyn, The Meaning of the Separation of Powers (1965: New Orleans: Tulane University Press, 1965) 39.
[19] [1984] USSC 181; 468 US 841, 846-47 (1984) quoting Nixon v Administrator of General Services [1977] USSC 147; 433 US 425, 468 (1977).
[20] A Dershowitz, ‘Comment: The Bounds of Legislative Specification: A Suggested Approach to the Bill of Attainder Clause’ (1962) 72 Yale Law Journal 350.
[21] L H Tribe, American Constitutional Law (Mineola: Foundation Press, 2nd ed, 1988) 631.
[22] T. Cooley, 1 Constitutional Limitations, vol 1 (Boston: Little, Brown & Co, 8th ed, 1927) 536-537.
[23] J Hatsell, Precedents of Proceedings in the House of Commons, vol 4 (London: Luke Hansard & Sons, 1818) 85-96, 100-102, 235-249, 323-346
[24] I May, Treatise on the Law, Privileges, Proceedings and Usage of Parliament, (London: Butterworths, 19th ed, 1976) 66.
[25] Ibid, 99.
[26]Halsbury’s Laws of England vol 34 4th ed, 522.
[27] Dershowitz, n 20 above, 331.
[28] [1810] USSC 10; 10 US 87, 136 (1810)
[29] United States v Brown [1965] USSC 134; 381 US 437, 488 (1965)
[30] n 6 above.
[31] Ibid 291
[32] [1915] HCA 58; (1915) 20 CLR 425
[33] Crimes Act 1915 ss 2 and 3
[34] n 32 above, 434.
[35] Ibid 436-437
[36] Ibid 441-442 (Isaacs J), 453 (Higgins J), 455-456 (Gavan Duffy and Rich JJ), 460 (Powers J).
[37] n 8 above, 539 (Mason CJ) 631 (Deane J), 689 (Toohey J), 706 (Gaudron J), 721 (McHugh J).
[38] Ibid 648.
[39] Ibid 647. Compare Mason CJ at 536, Toohey J at 685-686, and McHugh J at 721.
[40] Ibid 721
[41] Ibid
[42] Ibid 631.
[43] See for example International Covenant on Civil and Political Rights art 15 para 2
[44] n 8 above, 689.
[45] Art I § 9 cl 3 and Art I § 10 cl 1
[46] For an essay devoted to the argument that the ban on ex post facto law should be confined to `criminal' offences, see Raoul Berger, ‘Bills of attainder: a study of amendment by the Court’ (1978) 63 Cornell Law Review 355-404.
[47] n 12 above.
[48] Ibid 394
[49] Ibid 397
[50] Ibid 400
[51] For a list of cases in which the Court has stated that the attainder clause is confined to punishment for crimes, see A. Mueller, `Supreme Court's view as to what constitutes an ex post facto law prohibited by Federal Constitution' 53 L Ed 2d 1146 (Annotation). See also Collins v Youngblood [1990] USSC 113; 497 US 37, 42 (1990).
[52] [1866] USSC 23; 71 US 277 (1866)
[53] [1866] USSC 33; 71 US 333 (1866)
[54] n 52 above, 320
[55] [1946] USSC 104; 328 US 303 (1946).
[56] Ibid 315-316.
[57] n 29 above.
[58] n 28 above.
[59] [1878] USSC 29; 97 US 381, 385 (1878).
[60] P Winfield, Province of the Law of Tort (Cambridge: Cambridge University Press, 1931} 190.
[61] B L Benson, Enterprise of Law: Justice without the State (San Francisco Pacific Research Institute for Public Policy, 1990) 62.
[62] J W C Turner, Kenny’s Outlines of Criminal Law (Cambridge: Cambridge University Press, 19th ed 1996) 4.
[63] Blackstone, n 2 above, 54, 57.
[64] See, for example International Covenant on Civil and Political Rights Art 15(2) and European Convention on Human Rights Art 7(2).
[65] Proprietary Articles Trade Association v Attorney-General of Canada and others [1931] AC 310, 324.
[66] G Williams, Textbook of Criminal Law (London: Stevens, 1978) 14.
[67] Ibid 14-15.
[68] n 62 above, 5.
[69] M.P. Lehmann, `The Bill of Attainder Doctrine: A Survey of the Decisional Law', (1978) 5 Hastings Constitutional Law Quarterly 834.
[70] Tribe, n 21 above, 651
[71] United State v Brown, n 29 above, 458
[72] Cummings, n 52 above; Garland, n 53 above.
[73] Fletcher v Peck, n 28 above.
[74] Johannessen v United States 255 US 227 (1912).
[75] Burgess v Salmon, n 59 above..
[76] Hawker v New York [1898] USSC 67; 170 US 189 (1898).
[77]De Veau v Braisted [1960] USSC 76; 363 US 144 (1960).
[78] n 20 above.
[79] Ibid 356.
[80] n 29 above, 461.
[81] n 20 above, 352.
[82] n 29 above, 455.
[83] Wormuth develops a similar test in treating the imposition of disqualifications as inherently valid when there is no implicit censorial judgment of individuals, but only the derivation of presumptions about character from aspects of common knowledge or principles of general psychology. F D Wormuth, ‘Legislative Disqualifications as Bills of Attainder’, (1951) 4 Vanderbilt Law Review, 603, 610.
[84] Hawker v New York n 77 above.
[85] [1968] AC 717.
[86] Ibid 736.
[87].Ibid.
[88] n 20 above, 338.
[89] [1950] USSC 56; 339 US 382 (1949).
[90] n 29 above.
[91] Ibid 460.
[92] [1997] USSC 63; 521 US 346, 358 (1997)
[93] n 1 above, 98.
[94] Ibid.
[95] [1996] HCA 24; (1995-1996) 189 CLR 51
[96] Ibid 108, 109
[97] Attorney-General (Cth) v R; Ex parte Boilermakers’ Society of Australia [1956] HCA 10; (1956) 94 CLR 254.
[98] (1947-1948) 75 CLR 495
[99] Order made under regulation 14 of the National Security (Wheat Acquisition) Regulations authorised by the National Security Act 1939.
[100] Wheat Industry Stabilization Act 1946 s 11.
[101] n 99 above, 579.
[102] [1998] HCA 9; (1998) 193 CLR 173.
[103] (1995) 184 CLR 19.
[104] Ibid 188-189, 202, 210-211, 238, 275-276.
[105] (1973) 129 CLR 231.
[106] [1971] HCA 21; (1971) 122 CLR 114.
[107] n 110 above, 239.
[108] Ibid 231.
[109] Ibid 249.
[110] [1986] HCA 47; (1986) 161 CLR 88
[111] Ss. 143(2) and 118A(1)
[112] S. 143A(1) and (2)(a)
[113] Queen v Ludeke and others [1985] HCA 84; (1985) 159 CLR 636.
[114] n 108 above 94-95.
[115] Ibid 96-97.


AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback
URL: http://www.austlii.edu.au/au/journals/UQLRS/2007/1.html