1. Introduction
Suppose a law in a predecessor regime – call it ‘Rule X’ – required individualsFootnote 1 to report neighbours of a particular race, living within a specific vicinity of their residences, for extermination. The objective of Rule X was to support the predecessor regime's effort in carrying out genocidal acts. If individuals failed to comply with Rule X, the penalty was death. If they complied to avoid the sanction of a death sentence, should a tribunal in a successor regime involved in the process of transitional justice punish them?
If we accept Radbruch's thesis, as interpreted by Robert Alexy,Footnote 2 intolerably unjust laws are legally invalid. I refer to this position on the legal validity of intolerably unjust laws as ‘Radbruch's Formula’, which would render Rule X legally invalid. An individual who reported their neighbour was not justified by law. In substance, the individual abetted murder (or genocide) by assistance.
If this logical implication of Radbruch's Formula is accepted by a tribunal carrying out transitional justice,Footnote 3 at first sight, it does not seem fair for an individual who complied with the law. However, a law that requires the reporting of individuals of a certain race so that they could be put to death is morally heinous. From the perspective of the tribunal in the successor regime, if Rule X is not legally valid, criminal laws that were in force during the time of the predecessor regime, which prohibited abetment to murder, may be used to prosecute individuals who complied with Rule X, who were not justified by law as far as the tribunal in the successor regime was concerned. Some who reported their neighbours may have been racist and willing to comply, but some may have complied unwillingly. There are two conundrums that a tribunal in a successor regime may have to address in deciding whether to punish such individuals under criminal laws which were likely to have subsisted in the predecessor regime. The first conundrum is that the individual may have acted in accordance with what the individual believed was law. The second conundrum is that the individual may have acted under duress, out of fear of the penal sanction of death.
This article examines whether ignorance of law and duress should operate as defences for the acts done by individuals in compliance with intolerably unjust laws in a predecessor regime. I reason by comparison with how such defences are generally denied in relation to ordinary criminal acts (the ordinary criminal context), such as in the case of in individual claiming to be unaware that they are committing a crime (in relation to ignorance of law), or a terrorist coercing an individual at gunpoint to reveal the whereabouts of someone whom the terrorist wants to kill (in relation to duress). From this, I extrapolate to the extraordinary context of compliance with intolerably unjust laws (the Rule X context). I shall argue, by parity of reasoning, that they should not be regarded as defences in the context of transitional justice in relation to Rule X.
2. Radbruch's position on the legal invalidity of intolerably unjust laws
Radbruch applied his thesis to the Nazi regime.Footnote 4 He took the view that laws that are substantively unjust to an intolerable degree should not be regarded as legally valid, even when promulgated according to stipulated procedure.Footnote 5 This was particularly so when there was not even an attempt at justice, such as when equality was deliberately betrayed.Footnote 6 Alexy suggests that Radbruch gave due weight to the conflicting values of legal certainty and justice:Footnote 7 legal validity was lost only when the threshold of extreme injustice was reached. The need for certainty gives way to the demands of justice in the case of intolerable injustice. Radbruch said there were ‘principles of law’ that were ‘weightier than any legal enactment’, known as ‘natural law or ‘the law of reason’. While their details were ‘open to question’, ‘the work of centuries’ had ‘established a solid core of them’. There was ‘far-reaching consensus in the so-called declarations of human and civil rights that only the dogmatic sceptic could still entertain doubts about some of them’.Footnote 8 Thus, Alexy opines that judges should refer to a broad historical consensus to discern what constitutes intolerable injustice, circumscribing the judicial role in invalidating laws. Human rights violations involving jus cogens norms in international law can provide guidance by limiting instances that count as intolerable injustice,Footnote 9 such as those instances when laws are used to carry out acts of slavery and genocide. These are cases of intolerable injustice as there is clearly no attempt at equality.
I will not assess here the soundness of the relative weight that Radbruch and Alexy gave to the values of legal certainty and justice, as I have considered this in a recently published article.Footnote 10 My focus here is on the considerations before a court that gives Radbruch's Formula practical effect by holding that an intolerably unjust law is not legally valid. Radbruch's Formula operates differently in two time frames: during the rule of the regime in which such laws are found (the ‘unjust regime’) and after the unjust regime has been overthrown (at which point it will be referred to as the ‘predecessor regime’ and the succeeding regime will be referred to as the ‘successor regime’). The first part of this section considers the first benefit: during the rule of an unjust regime, a court is not required by law to punish an individual who valiantly refused to comply with an intolerably unjust law because that ‘law’ is not legally valid. The second part of this section considers the scenario after an unjust regime has been overthrown: an individual who complied with an intolerably unjust law in the predecessor regime may be found by a tribunal in the successor regime in carrying out transitional justice to have acted criminally if the act would have been criminal under the then-subsisting criminal law but for the legal justification provided by the intolerably unjust ‘law’. The rest of this article focuses on the second part, the time frame during which the implication of Radbruch's Formula would be relevant. However, I explain the operation of Radbruch's Formula in the time frame of the first part to emphasise how the same problems – the two conundrums – are unlikely to arise.
2.1. During an unjust regime
A legal positivist such as Hart, who rejects Radbruch's Formula, would regard an intolerably unjust law as legally valid as long as the criteria in the regime's rule of recognitionFootnote 11 are satisfied. During an unjust regime, judges hearing cases in which individuals have valiantly disobeyed an intolerably unjust law may choose to stay true to their moral obligation not to punish individuals who do what is morally right and disobey the law. Legal positivists recognise this tension between the legal and the moral: Hart notes that Austin and Bentham believed that there would be a plain moral obligation to resist laws that reached a certain degree of iniquity, but legal positivists maintain there remains a legal obligation.Footnote 12
In contrast, if Radbruch's Formula is affirmed, intolerably unjust laws are legally invalid. Conceptually, Radbruch's Formula gives judges a legal justification to strike down an intolerably unjust law. In practice, judges may be reluctant to invoke Radbruch's Formula while the unjust regime prevails out of fear of repercussion. Consider Rule X and the case of an individual prosecuted for failing to report the neighbour. According to Radbruch's Formula, the individual should not be punished; the Formula offers better protection for individuals wishing to do right. Judges can serve as counter-majoritarian checks against majorities in the populace working through elected legislatures to enact intolerably unjust laws in very limited instances, particularly when creative statutory interpretation solutions to get around intolerably unjust laws are not available. While it can sometimes be difficult to assess whether injustice to an intolerable degree is present, the mere risk of overreach in some scenarios provides no good reason to reject judicial action in all cases. There may also be cases where individuals complied with Rule X. Applying Radbruch's Formula, Rule X is not legally valid. Their acts lack the justification of law; their acts amount to abetting genocide or murder.Footnote 13 In theory, such individuals can be punished under subsisting criminal laws for abetting the crime of murder.Footnote 14 Prosecution is unlikely in practice if the executive branch of government is ad idem with the legislative branch. During the rule of an unjust regime, Radbruch's Formula is more likely to be invoked to exonerate those who refused to comply than to prosecute those who did comply. Since it is when individuals who complied with intolerably unjust laws are prosecuted that the two conundrums arise, the two conundrums are more likely to arise in a transitional justice situation. We shall therefore focus on the transitional justice context.
Before moving on to consider this in the transitional justice context in Section 2.2, however, it bears mentioning that interpretive approachesFootnote 15 apart from Radbruch's Formula may be employed by good judges in unjust regimes in seeking to interpret such laws to achieve outcomes that serve justice. Dyzenhaus, for instance, notes that the resort to purposive approaches rather than supra-positive laws might be possible in some cases.Footnote 16 If so, a good judge in an unjust regime who engages in a creative interpretation of the law might exonerate a valiant individual who refuses to comply with an intolerably unjust law, without having to resort to Radbruch's Formula. The focus of this article, however, is on transitional justice by a tribunal in a successor regime. We are therefore concerned with cases where individuals complied with intolerably unjust ‘laws’ and had already committed heinous acts out of fear of being punished for non-compliance in the predecessor regime.
2.2. Transitional justice by a tribunal in a successor regime
According to Radbruch's Formula, Rule X was not legally valid in the predecessor regime. A tribunal effecting transitional justice in a successor regime may consider using the subsisting criminal law to punish individuals who complied with Rule X for abetment to murder or genocide. Punishment is not retroactive in substance because Rule X was not law during the predecessor regime's rule, which means that someone who reported their neighbour did not have legal justification for doing so. Thus, a then-subsisting criminal law provision, such as abetment to murder by assistance, could be resorted to in order to prosecute.
Rejecting Radbruch's Formula would mean that punishing individuals who complied must be undertaken through retroactive laws.Footnote 17 The act was properly justified by law when committed. Retroactive punishment is generally regarded as unfair, and punishment might not be pursued. As noted by HartFootnote 18 in his debate with Fuller,Footnote 19 there is a moral dilemma: the good to be served by punishment must be considered alongside the unfairness of retroactive laws.
Sometimes, subsisting criminal laws may not be appropriate or available for punishing individuals. Elements of existing offences may not be satisfied,Footnote 20 or it might be argued that abetment by genocidaires should not be trivialised by treating it as abetment to murder. However, Radbruch's Formula remains conceptually important. Individuals of the state have participated in acts of intolerable injustice, which are not justified by law. Prosecution can be supported by customary international law norms today, or by the formulation of a crime against humanity (as was done in the Tokyo and Nuremberg tribunals in the 1940s),Footnote 21 or common law criminal law, which arguably can be resorted to as extremely heinous acts have been committed.
If punishment is pursued as a matter of transitional justice after the unjust regime has been replaced, the two conundrums arise. I shall argue that widespread endorsement of Radbruch's Formula, alongside well-known instances of historical atrocities such as those perpetrated by the Nazi regime, render the first conundrum – stemming from the individual's claim that they believed they were complying with law – of less concern because that claim becomes less tenable today. As for individuals acting under duress during an unjust regime, some individuals would be torn between their moral obligation to refrain from heinous acts and their concern that judges in the unjust regime might not apply Radbruch's Formula and might punish them for failure to comply. Adding the threat of future punishment as a matter of transitional justice might in fact incentivise them to do right. The result is that even the most pragmatic individuals, who are otherwise willing to put aside their moral sense, might resist such laws out of fear of being punished in future as a matter of transitional justice. It would help to ensure that perpetrators of heinous injustice cannot realistically use the instrument of law to carry out their objectives. Atrocities would have to be committed outside the law, through brute force, laying bare their unjustifiable nature. Individuals and the legal profession might resist intolerably unjust laws, or pockets of civil resistance might arise against the brute force of the government if atrocities are committed outside the law. While Radbruch's Formula cannot prevent unjust regimes from coming about, its supporters are not necessarily guilty of the naivety of which Hart accuses them when he observes that the positivistic slogan acquired a sinister character only in Nazi Germany, but elsewhere went with the most enlightened attitudes.Footnote 22 One counts on the resistance towards intolerably unjust laws at the very onset of such unjust regimes if Radbruch's Formula is widely endorsed, such that we would not be willing to countenance the validity of intolerably unjust laws.
I should add that Radbruch himself might have been inclined to grant a concession to individuals who complied. In considering judicial practice towards ‘statutory lawlessness’ (that is, intolerably unjust laws), Radbruch examines whether judges are culpable of homicide in sentencing individuals under such laws. While he thinks that judges should serve justice at the price of their own lives, Radbruch notes that they might plead ‘necessity’ if subsequently prosecuted, given that they ‘would have risked their own lives had they pronounced National Socialist law to be statutory lawlessness’.Footnote 23 In view of how judges are given a concession, Radbruch might have approached individuals who complied out of fear for their own lives even more leniently. However, the objective of this article is not to find out what Radbruch's view might have been on whether individuals who complied ought to be punished. Instead, I explore the conceptual implication of intolerably unjust laws being legally invalid.
3. Historical instances
Before turning to the two conundrums to be confronted by a tribunal in a successor regime implementing transitional justice, I consider two well-known historical instances: one in relation to which Radbruch's Formula was formulated – the Nazi regime; the second in which the Formula was invoked by some – the cases involving the East German border guards. I explain why I choose to examine the fairness of punishment as a matter of transitional justice in the context of my hypothetical scenario of Rule X, rather than in these historical cases.
The ‘Case of the Grudge Informer’ is considered by Hart and Fuller in their debate over the connection of law with morality in the context of Nazi laws,Footnote 24 and has been reconsidered by others.Footnote 25 There was no duty under the statute to report the persons who had criticised the government, and a wife had reported her husband to get rid of him.Footnote 26 The Case of the Grudge Informer is not useful for the purpose of this article as I am concerned with the fairness of punishing individuals who were compelled to comply with intolerably unjust laws and, in that sense, forced to participate in the heinous acts of a predecessor regime.
The cases involving the prosecution and trial of the East German border guards – soldiers in the German Democratic Republic (GDR) who shot and killed East German citizens who tried to escape across the wall separating East and West Germany prior to their reunification – have been extensively discussed. Under the pertinent statute in the GDR border guards may, if necessary, shoot a person to prevent or stop the commission of a major crime.Footnote 27 Various soldiers were prosecuted for homicide or murder on the ground that they lacked legal justification for their acts. This could be because the statute relied upon did not justify the acts when properly interpreted. Alternatively, if the GDR authorities interpreted those acts as justifiable, they were not legally justifiable after all, for example, because the statute violated international human rights standards according to which everyone should be free to leave any country,Footnote 28 or it violated Radbruch's Formula.Footnote 29 It is questionable whether the Formula applies in such a scenario, which differed from the atrocities committed in the Nazi regime.Footnote 30 Unlike the Nazi regime, there was no attempt in the GDR cases to wipe out an entire race.Footnote 31 The relevant law could be interpreted as an attempt to protect the integrity of a country's border. Even if the statute did authorise (or require) the shooting, it was not undoubtedly an intolerably unjust law of the nature with which Radbruch had been concerned. In this article I have used the hypothetical Rule X to consider the most obvious instance of an ordinary individual of the law confronting the requirement of an intolerably unjust law to examine whether it would be fair to punish such an individual. In an obvious case of intolerable injustice, it becomes more tenable to suggest that the individual ought to have known better – that what purports to be law is not actually law. Furthermore, examining the GDR cases adds layers of complication as to whether they are in fact authorised to shoot under the statute in the circumstances.Footnote 32 A hypothetical Rule X scenario enables us to home in on the conundrums, to which I now turn, as my answer upon such examination is that it is not unfair to punish such individuals.
4. First conundrum: Belief as to legality of act
The first conundrum in punishing individuals who complied with Rule X is that they may have believed that Rule X was law. Individuals who have complied with intolerably unjust laws may argue that they should not be faulted for thinking they had a legal obligation. How can an individual be expected to know that Rule X was not law according to Radbruch's Formula? The Formula represents one out of several possible positions on legal validity. Rule X is contained in a statute passed by the legislature. Individuals may readily regard that as legally valid.
However, every court or tribunal faced with the enforcement of a law is tasked firstly with deciding on its legal validity. If the court adopts Radbruch's Formula and decides Rule X is not legally valid, but the individual had thought Rule X to be valid law, this seems to be a case of ignorance of the law on the part of the individual, which, albeit controversially, would not generally be an excuse under (Anglo-American) criminal lawFootnote 33 if this were an ordinary criminal case.
Two main issues arise for consideration. First, and preliminarily, is a belief as to the legal validity of Rule X, ‘mistaken’ from the point of view of Radbruch's Formula, truly like a case in which the individual acts not knowing the law in the ordinary criminal context? We must find crucial similarities before we can proceed to argue, by parity of reasoning, from how ignorance of the law is treated in the ordinary criminal context to how it should be treated in the Rule X context. Second, why is ignorance of the law generally not an excuse in the ordinary criminal context and, by parity of reasoning, do those reasons apply in the Rule X context?
As for the first and preliminary issue, it can be argued that this is not an ordinary case of ignorance of the law: the legal validity of Rule X turns on a theoretical debate as to what law is. The content of Rule X is in fact fully known; it is the fact that it is not law (according to Radbruch's Formula) that is not known. Individuals cannot be expected to resolve a theoretical debate between legal positivism and natural law theory in favour of natural law when officials, judges and, indeed, legislators fail to agree. Legal positivists such as Hart regard intolerably unjust laws as laws if the rule of recognition stipulates for their legal validity.Footnote 34 Furthermore, the natural law position on lex injusta non est lex is not clear. While Hart, a positivist, regards natural law theory as taking the view that unjust laws are not law,Footnote 35 Finnis regards it as a subordinate theorem.Footnote 36 Radbruch's Formula has been referred to as a non-positivisticFootnote 37 rather than a straightforwardly ‘natural law’ position.Footnote 38
A theoretical debate over the definition of law, on its own, however, is not a good enough reason for excusing individuals who complied with Rule X while claiming ignorance of the law. A theoretical debate always exists, even when it is not in focus because there happens to be no disagreement as to whether a particular law is legally valid across the spectrum of possible theoretical positions on legal validity.
I therefore turn to the second issue: whether the arguments for and against ignorance of the law being an excuse in the ordinary criminal context apply, by parity of reasoning, to the Rule X context. Suppose individuals argue that they should not be punished because they believed they were complying with the law or did not know that an intolerably unjust law was not law. I shall consider the following. Do arguments against the defence of ignorance of the law in the ordinary criminal context apply equally to the scenario of active reliance on Rule X, or, instead, do arguments to allow the defence in specific cases in the ordinary criminal context apply to the Rule X context?
4.1. Why the defence of ignorance of the law is generally not allowed
Generally, Anglo-American systems of law endorse the notion that ignorantia legis non excusat Footnote 39 in the ordinary criminal context. Allowing the defence of ignorance of the law would favour the ‘most obtuse and stolid of criminals’.Footnote 40 Selden's concern, for example, is that it would be too easily pleaded and hard to confute. However, this concern can be addressed by placing the burden on those pleading the defence to show they actually do not know the law.Footnote 41 Indeed, some existing laws (tax laws, for example) include a requirement of wilful violation or allow a defence of reasonable mistake, which raise similar difficultiesFootnote 42 and are not regarded as insurmountable. The courts in South Africa have not been plagued by practical difficulties of every person pleading the defence after the defence of ignorance of law was made available.Footnote 43 Importantly, it is also unlikely that many would argue that they do not know that ‘it is illegal to murder, rape, rob, burgle, steal, cheat, lie’,Footnote 44 such claims being rather far-fetched.
Whenever one can fairly presume knowledge of the law, there is good reason to reject the defence of ignorance of the law.Footnote 45 At a very general level, the requirements of Fuller's rule of law – such as that laws are clear and are publicised or made known – suggest that it is important that those governed (and punished) by law know its content.Footnote 46 There is no dissonance between affirming Fuller's requirements that emphasise knowledge of the law and affirming the idea that ignorance of the law is no excuse. If the legislator fulfils Fuller's requirements by making and disseminating laws that are clear, the legislator has done what it needs to do for the rule of law. Generally, legal systems cannot be expected to go further to ensure that individuals actually know the law. Thus, at first sight, it seems fair to presume knowledge if Fuller's requirements have been fulfilled.
However, presuming knowledge in the past might have been reasonable when it was relatively easy to acquire knowledge of the law, as there were not many offences.Footnote 47 It is problematic today; there are many more offences,Footnote 48 particularly those that are perceived to be regulatory.Footnote 49 Furthermore, sometimes (such as in the case of English law today) the content of criminal law is not readily known, being found in many statutory instruments as well as the common law.
That said, even if we do not presume knowledge of complex areas of the law, it is reasonable to impose a moderate duty on individuals to know the law; such a moderate duty permits a ‘general but circumscribed defence’ of ignorance of the law.Footnote 50 However, individuals cannot know the law in detail, given its complexities and how legal training is required in order to understand the scope of laws. If there is only a moderate duty on individuals to know the law, it does not seem fair to hold them liable when they are not familiar with the details and complexities. Why, then, do we deny individuals the defence of ignorance of law in such cases,Footnote 51 particularly when they are neither blameworthy nor negligent, as they have fulfilled their moderate duty to know the law?Footnote 52
The reason is that the demand to treat individuals fairly and permit the defence is outweighed by other interests: allowing the defence of ignorance of the law would encourage ignorance precisely where the lawmaker wants individuals to know the law.Footnote 53 Notably, this is a consequentialist or utilitarian consideration. It is at odds with the retributivism to which many legal philosophers subscribe in relation to criminal law if punishment is in fact undeserved because of ignorance of the law.Footnote 54
Aside from presuming knowledge of the law, whenever the law has tracked moralityFootnote 55 – such as when the common law was a source of English criminal law and addressed what was morally wrong according to community mores or long-established ideas of right and wrongFootnote 56 – it may not be unfair to find one liable without specific knowledge of the law. The reason is that there is moral culpability in the commission of the acts prohibited by such laws. The counter-argument to grounding liability in moral culpability, however, is that it is not appropriate or fair to punish someone for failing to behave in a manner that is morally salutary. Individuals may choose to obey the law with the motivation only to avoid legal liability: their ignorance, at least in cases where they are blameless for not knowing the law, ought to constitute an excuse.
In any case, grounding criminal liability in moral culpability applies only to laws that track morality. For this reason, some scholars distinguish between the treatment of mala prohibita and mala in se offences.Footnote 57 For mala prohibita offences, even a ‘good’ person might make mistakes as to the content of the law. It is the violator's knowledge of illegality that makes their act morally culpable. For mala in se offences, a person can be regarded as culpable for doing what is morally wrong, as embodied in the law.Footnote 58 A person who is ‘procedurally and epistemically faultless’ – that is, ‘not ignorant as a result of recklessness, negligence, or other forms of belief mismanagement’ – may be blameless in legal ignorance for mala in se offences but demonstrates moral ignorance, which constitutes or reveals a morally objectionable failure to care about what is morally significant.Footnote 59 Liability for mala in se offences can thus be seen as grounded in the very moral culpability in committing the act. Committing the act, in itself, says something about the ‘moral quality of (the actor's) values, motivations, and emotions’.Footnote 60 It should not matter that the person is unaware of the illegality of the act. In grounding criminal liability in moral culpability, the defendant is held to standards of morality that every person must be taken to know. This is disputable in cases concerning specific moral norms over which there is no moral consensus,Footnote 61 and even for mala in se offences, in relation to ‘outer limits of the law’,Footnote 62 it would be fair to require knowledge of illegality or some form of fault. Serious crimes, such as sexual offences, are mala in se but changes in the law may concern the ‘outer limits of the law’: even if most must know there are laws relating to sexual offences, they may be unaware of whether a particular activity is within law's ambit.Footnote 63 The existence of a ‘moral core’ of ‘real crime’ does not make such decisions clearer at the outer limits.Footnote 64
In the ordinary criminal context, if an individual abets by assisting in a race-motivated murder and pleads that they did not know their act violated the law, this plea is unlikely to be well received. This clearly involves a malum in se offence. There is strong moral consensus that the crime is egregious.Footnote 65 If ignorance were pleaded in relation to crimes that enforce more contentious moral norms, mala prohibita offences and crimes at the ‘outer limit’ of the law in mala in se offences, questions of an appropriate ground of liability and of fairness might arise. For abetment to race-motivated murder, an individual who pleads ignorance of law has such failure of character that few would argue that it is unjust to punish that person. This crime is in fact analogous in substance to the act mandated by Rule X. There are strong reasons to reject the plea of ignorance of the law, that is, ignorance that Rule X is legally invalid, and that the act of reporting the neighbour is legally unjustified and constitutes abetment by assisting in a race-motivated murder.
However, a truly analogous situation (to Rule X) in the ordinary criminal context is more nuanced. The analogue of the Rule X case in the ordinary criminal context is the case of a person who is threatened with death if they do not assist in a race-motivated murder. A person who assists in the ordinary criminal context can be prosecuted for abetment to murder by assistance. Such a person would also try to plead the defence of duress (considered in Section 5). Suppose we find that duress is rightfully not a defence to abetment by assisting in murder in the ordinary criminal context (as will be argued in Section 5). Can such a person in the ordinary criminal context run a more complex defence of ignorance of the law? Can they argue that they are ignorant of the law in the sense that they thought that duress is a defence in a case of abetment to murder by assistance? The ignorance seems, at first blush, less abhorrent than ignorance of the law against abetment to murder by assistance. This is particularly so in view of the debate over whether the defence of duress should be available for murder. The right solution might depend on whether we think the duress defence should be available in the case of abetment to murder. If it should not, and to the extent that denial of the defence and consequent illegality of the act are based on its moral blameworthiness, this suffices to ground liability, based on moral culpability, in the ordinary criminal context. The same should apply in the Rule X context.
4.2. The special case of legitimate reliance
There remains a difference, which some might find significant, between the ordinary criminal context and the Rule X context. In the ordinary criminal context, an individual usually claims not to know a law that makes the conduct illegal. In the Rule X context, however, the individual actively relies upon an enacted law, Rule X, which turns out to be legally invalid according to Radbruch's Formula.Footnote 66 What difference should active reliance on a purported law make to the arguments for denying the defence of ignorance of the law? There are some analogous situations in the ordinary criminal context from which we can draw relevant principles.
In the ordinary criminal context, where an individual reasonablyFootnote 67 relies on official advice as to the law, it has been argued that an estoppel should apply against the authority in suggesting that ‘ignorance of the law is no excuse’ when it subsequently takes the view that the advice is misguided.Footnote 68 From a deontic point of view, the estoppel honours the legitimate expectation of the individual who is seeking to be law-abiding and to comply with the law. Permitting a defence in the case of officially induced error also encourages citizens to check on the lawfulness of proposed activities.Footnote 69 From a consequentialist point of view, it promotes legal knowledge and compliance.Footnote 70
Aside from official guidance, individuals may also be guided by a court decision (prior to its being set aside by a higher court),Footnote 71 as it would be unfair to expect individuals to know the law if even a court was ‘ignorant’.Footnote 72 Individuals may also be guided by statutes before they are held to be unconstitutional.Footnote 73 Court decisions and statutes are more authoritative than official advice: the court is the final authority in determining questions of law; the legislator is tasked with making law.
A final point worth emphasising is that individuals in such circumstances may reasonably rely on official advice as to the law, court decisions and statutes; thus, the authority is estopped from insisting that ignorance of the law is no excuse. An argument can be made that in rare circumstances it would be unreasonable for such reliance to be made, or that individuals should not be allowed to raise estoppel. For example, an individual who bribes an official to provide official advice favourable to the individual does not come with clean hands and should not be allowed to raise estoppel.
The Rule X context, involving a statutory enactment, is similar to the case of reliance on a statute before it is held to be unconstitutional. Despite the even stronger case for justified reliance on this over official advice, Rule X is no ordinary circumstance. After the Nazi regime and with recognition of genocide as an international crime, it is unreasonable for an individual to rely on Rule X claiming that they could not have known better. It would not be unfair to deny estoppel based on reliance on a purported law such as Rule X. Radbruch's Formula, applying to intolerable injustice, elevates the concerns of justice at the smallest possible sacrifice to legal certainty. However, where there is genuine dispute as to whether the laws are intolerably unjust,Footnote 74 an individual who acted according to their conscience should be allowed to avoid liability on the basis that they actively relied on the law. They should not be punished by the tribunal in carrying out transitional justice. Likewise, individuals in entirely closed political regimes who are unconnected with the world and do not know about genocide being criminal, and who have been so brainwashed as to lack the capacity to realise the egregious nature of the act, lack moral blameworthiness. Such persons may be permitted a limited plea of ignorance of the law in the Rule X context.
5. Second conundrum: Acting under duress
Individuals who complied with intolerably unjust laws may have done so out of fear of punishment. If the sanction was relatively minor compared with the sanction their neighbours faced for being of a particular race, punishing such individuals is less objectionable, but what if the sanction for disobedience was death? A second conundrum arises over the fact that an individual might have acted under duress. Should the individual be excused, particularly if the individual risked the sanction of death for disobedience?
Consider an analogous situation in the ordinary criminal context: one person kills another on the command of a terrorist who threatens the first with death if they do not do so. Such a scenario in the ordinary criminal context has been considered by courts, law commissions and academics.
Before examining the defence of duress in the ordinary criminal context, I would note that this defence was also denied to those who complied with (impliedly coercive) superior orders to kill in the context of Nuremberg. The Nuremberg context was considered when the defence of duress was denied in a murder charge in criminal law.Footnote 75 Intolerably unjust laws that apply to ordinary individuals in the Rule X context raise different concerns from heinous superior orders. However, given how the Nuremberg context is used to argue for the position with regard to duress in murder charges in the ordinary criminal context, one might think that the question of rightful punishment seems to be regarded as more intuitively justified in the Nuremberg context than in the ordinary criminal context. Instances involving superior orders are akin to criminal law scenarios in which defendants have voluntarily placed themselves in the situations that led to duress. The defence has been denied in such situations, such as when the defendant joined a criminal gang or a terrorist organisation.Footnote 76 That said, differences in individual cases should be considered. For instance, an official who joins a military unit, serving in a regime which subsequently orders morally heinous acts does not intend at the outset to undertake such activities. It may be pertinent for moral accountability to assess if the official had the opportunity to resign. The Nuremberg context offers a less perfect analogy with the ordinary criminal context in which the person under duress is not at fault for being placed under duress. What about the Rule X context?
The Rule X context shares more similarities with the ordinary criminal context (in which the person under duress is not at fault for being placed under duress) than with the Nuremberg context of heinous superior orders. Rule X simply applies to some individuals upon its being enacted, without their having done anything to attract its application. Also, criminal law duress in general has received much attention in recent decades; we can therefore draw upon insights from criminal law duress for the Rule X context. I examine the treatment of the duress defence in the ordinary criminal context in relation to murder and what it might suggest, by parity of reasoning, for the Rule X context.
5.1. Sanctity of life versus impairment of will
Whether duress should be a defence to murder or abetment to murder is much debated. Under English common lawFootnote 77 duress is generally not a defence to murderFootnote 78 because of the belief in the sanctity of life:Footnote 79 it is not justifiable to take another life to preserve one's own as comparative lives cannot be weighed.Footnote 80 Exceptionally, some states in the United States allow the defence of duress for a felony murder charge. An accessory to murder was allowed the defence in the United Kingdom, although the position has since changed.Footnote 81 In so far as a moral dilemma is presented to the person threatened, it has been argued that the test of proportionality – comparing the threatened harm with the moral evil the person is compelled to do – should be used to grant some form of concession to the person threatened where punishment is concerned.Footnote 82
In the Rule X context, the concern for sanctity of life is the same as it is in the ordinary criminal context, but the problematic assessment of proportionality is seen starkly in the Rule X context. It can come down to comparing the size of one's household with that of one's neighbour: an individual who reports their neighbour, as required by Rule X, may do so if they hope to save their family of five but their neighbour is single; an individual who reports their neighbours, as required by Rule X, may not do so if they are single while their neighbours are a large family of five.
A dominant underlying rationale for the defence in the ordinary criminal context is that duress impairs (even if it does not overbear) one's free will, rendering punishment unjust.Footnote 83 The act is voluntary in that it is volitional, but there is moral non-voluntariness.Footnote 84 A person's choice among limited alternatives has been ‘rigged by an undeniably culpable third party’.Footnote 85 If one's act is not voluntary in some sense, one may not deserve a conviction, or should be entitled to mitigation.Footnote 86 However, denying the defence and refusing to consider duress in mitigation might enhance the deterrent effect when one is strongly tempted to commit a crime,Footnote 87 though this is questionable if a reasonable person could not resist the threat.Footnote 88
The less immediate nature of the threat in the Rule X context weighs against affording the defence. The individual compelled by law to report their neighbour does not face a threat to their life at gunpoint. The individual can contemplate alternatives such as escaping, advising the neighbour to escape, or stalling for time. Assuming the processes of law are followed through, a non-reporting individual must be discovered for failure to report their neighbour, prosecuted and convicted, before the sentence is carried out. Even if the individual fears that the processes of law would not be properly adhered to in a rogue state and therefore reports their neighbour, the act is neither instinctive nor unthinking; nor is the will similarly impaired. This weighs against the defence being available in the Rule X context.
What about the peculiar power of a heinous state to oppress through inciting fear? Montesquieu, for example, describes the paralysing fear (crainte) that a despotic government can create. This leads individuals to have an instinct for self-preservation that ‘keeps the individuals in an animal-like obedience’.Footnote 89 It leads to a ‘permanent state of foreboding’ as opposed to being a ‘sudden response to danger’ (peur).Footnote 90 Montesquieu's claims require investigation by social psychologists. Heroism among ordinary people who pay the price with their own lives to save their neighbours in oppressive regimes are well documented, though Montesquieu does not rule this out. Possibly, while the threat is not as immediate as it is in the ordinary criminal context, the atmosphere of fear substantially overpowers the will such that it may not be just to carry Radbruch's Formula to its logical conclusion and effect punishment on those who complied.Footnote 91 Whether there is such overwhelming fear may also depend on how advanced the despotic regime is.
Supporters of the duress defence in the ordinary criminal context argue that regarding duress as an excuse does not derogate from the commitment to the worth of human life. It merely acknowledges that acting out of self-preservation is ‘understandable’.Footnote 92 While it arguably ‘(mollifies) ethical weaknesses’,Footnote 93 criminal law should not punish individuals for not attaining the pinnacle of ethical behaviour,Footnote 94 or demand heroism of the threatened person,Footnote 95 when a person of reasonable moral strength cannot fairly be expected to resist the threat.Footnote 96 The American Law Institute's Model Penal Code, for example, asks the jury to consider whether a person of ‘reasonable firmness’ would be able to resist the threat. This seems to require ‘a moral judgment regarding the ambit of fortitude of individuals constrained by prevailing circumstances’.Footnote 97 Yet taking reference from dominant or likely social practices without considering the moral value of such practices seems contrary to how the law often sets a standard regardless of dominant social practice.Footnote 98 While Lord Hailsham in R v Howe has been criticised for expecting heroism of the ordinary person, and refusing to protect ‘the coward and the poltroon’,Footnote 99 innocent victims deserve protection too from those seeking to kill them for self-preservation.Footnote 100 Even if the ordinary person might not have been deterred by future punishment, deterrence is not the only reason for penal laws: the law can affirm moral values such as sanctity of life for a moral educative function.Footnote 101
A theory of duress is necessarily ‘moralised’ and normative even if it appears on its face to focus on psychological statesFootnote 102 or what an average person would do. Those who reject the defence treat sanctity of life as non-derogable. Those who argue for the defence on moral grounds suggest that if law is connected with morality, unjust punishment should be avoided.Footnote 103 In the Rule X context, regarding the intolerably unjust law as legally invalid means that the act of reporting one's neighbour is legally unjustified; it is abetment to genocide or murder. Given that Radbruch's reason for treating intolerably unjust laws as legally invalid is to give justice its due place, it can be argued that it would be ironic if unjust punishment results. However, it should also be noted that a theory of duress is normative: a decision not to excuse cowardice in cases when one's will is merely impaired but not overborne sets a standard of behaviour and is not necessarily unjust as the neighbour's life is also at stake.
5.2. Consequentialist rationales for the defence of duress
The duress defence in the ordinary criminal context has been supported by consequentialist rationales. If a threatened person regards the harm with which he is threatened as the greater evil when compared with the harm that he does,Footnote 104 then, adopting Hobbes's view, Austin and Bentham suggest that the person, confronted with the choice between two evils, should not be faulted for preferring self-preservation.Footnote 105 Moreover, in terms of Pareto-optimality, no one would be better off in some scenarios, as all would die anyway if the person threatened does not carry out the wishes of the person issuing the threat.Footnote 106 However, assessment of Pareto-optimality in the Rule X context is more complex, depending on factors such as the efficacy of the heinous state and the likelihood of other individuals reporting those of the targeted group. Even if the heinous state is highly efficacious, Pareto-optimality may not count in favour of affording the duress defence in the Rule X context, as there is a better chance of both the coerced and the innocent victim escaping enforcement of the unjust law.
Another version of the consequentialist defence in the ordinary criminal context compares the benefit of punishing the person threatened with the benefit in not punishing. The fear of the threatened harm may be substantial and imminent enough such that no deterrent value is posed by the threat of future punishment. Also, punishment does not serve the rehabilitation of such an offender.Footnote 107 Deterrence of other potential offenders in the form of genocidaires has been thought to be ineffective, as some commit genocide precisely after assessing the benefits of acting. Moreover, the uncertainty of punishment by international criminal law, historically, results in failure of deterrence.Footnote 108
In the Rule X context, the non-instantaneous nature of the threat, compared with the ordinary criminal context, increases the chance of a deterrent effect. Such effect should be assessed over all individuals in totality, rather than individually. There is also a moral educative effect which underlies the punishment of such crimes under international criminal law as a matter of transitional justice. Punishment sends the message that laws must be minimally just. With some individuals deterred, a despot who wishes to execute heinous acts may have to do so outside the law. The widespread endorsement of Radbruch's Formula – of the idea that intolerably unjust laws are not laws – increases the likelihood of resistance to despots working through law from the outset, not just from the legal profession or those trained in the law but from officials and individuals. Ideally, the rise of despots by harnessing law to effect their cruel acts is arrested at the outset: the widespread civilian and official endorsement of Radbruch's Formula ironically and happily renders its invocation for the sake of transitional justice unnecessary in future. The case to deny the defence of duress on consequentialist grounds is strong in the Rule X context as there are benefits peculiar to the objective of transitional justice. The benefits extend beyond deterring persons from acting out of self-preservation to conveying that laws must be just, and legal authority must respect dignity. Generally refusing to treat civilians as helpless might empower them to stand up to the onset of rogue regimes and prevent their advancement. The defence of duress should not be allowed.
5.3. Does it make sense to reject the defence of duress in the light of other aspects of criminal law?
Is rejecting the duress defence in the ordinary criminal context consistent in principle with other aspects of criminal law? I argue that it is.
In so far as a person acting under duress makes a rational and intentional choice to preserve their life, it is arguably consistent in principle to reject the defence given that ‘malice aforethought’ has been rejected as a requirement of mens rea.Footnote 109
Allowing the defence in other cases without at least a partial defence to murderFootnote 110 is explicable as sanctity of life cannot be derogated from for the sake of self-preservation where the victim is innocent. Short of this, the law permits one person to use another as a means.Footnote 111
There is also principled consistency despite how sanctity of life is regarded under other defences, such as self-defence. While the instinct of self-preservation is strong and natural in cases of duress and self-defence,Footnote 112 a relevant distinguishing factor from self-defence is the innocence of the person whose life is being violated. Moreover, if the doctrine of double effect is used to analyse duress in such a scenario, death is intended; one's being let off is a separate act that follows upon one's act of causing death intentionally.Footnote 113 In cases of self-defence, the force used has to be proportionate before one can rely upon the defence, and there is no intention to cause death – only a side effect of death. If denying the duress defence for murder in the ordinary criminal context stems from respect for sanctity of life, which cannot be derogated from for the sake of self-preservation, the pressing need in the Rule X context to ensure that despots cannot use civilians en masse to effect their purposes against other innocent individuals is all the more important.
5.4. De facto recognition of an alternative authority
The defence of duress in the ordinary criminal context presents a jurisprudential puzzle. Austin views law as orders backed by threats of sanction. Absolving the person threatened suggests that individuals can avoid legal consequences by coming under what is de facto an alternative authority that uses coercive force. This problem was put into sharp focus in cases of duress by terrorists from the Irish Republican Army (IRA). Allowing the duress defence makes it seem as if the IRA's coercive order can confer immunity on the defendant.Footnote 114 What would have been an unlawful act carried out by terrorists (the IRA) can be transformed into a ‘lawful’ act by the murderer who acts under duress, as the murderer is not punished if the duress defence is allowed. The deterrent effect of criminal law is rendered nugatory where it is most needed. The coercive power of the state exercised through law is overwhelmed by another power.Footnote 115 In cases in which terrorists and others use ordinary individuals to achieve criminal ends, criminal liability has been viewed as notionally important. Duress should be countenanced, if at all, only in mitigation.Footnote 116
In the Rule X context, the source of coercion is a state acting through a purported law, rather than terrorists. This presents no jurisprudential puzzle from the legal positivist's perspective. Particularly for Austin, Rule X is straightforwardly law. From a non-positivistic perspective, resistance towards the rogue state, rather than compliance with its orders, is at least as important as it is for the terrorist scenario. Criminal liability may be helpful to incentivise resistance.
It has been suggested, in the ordinary criminal context, that the duress defence should be allowed when a person acts under duress from terrorists and similar, as criminal law should not be blind to the relatively lower degree of moral blameworthiness of those acting under duress. We should not be concerned about terrorists using ordinary individuals to carry out acts of violenceFootnote 117 as that is rare. Moreover, the law still responds – by punishing those who exerted the duress,Footnote 118 and defendants who exposed themselves voluntarily to such threats can be denied the defence.Footnote 119
In contrast, though, in the Rule X context, state perpetrators intentionally get individuals on board through ‘legal’ coercion as it enables such regimes to gain traction. It is more compelling as a general rule not to allow the defence of duress.
Should it make any difference that the ‘highest’ power in the territory is putting individuals under duress? This is unlike the situation of terrorists putting individuals under duress, when terrorists operate under another legal power. Hale examines the non-availability of the duress defence for murder in peacetime separately from times of war, public insurrection or rebellion, when a different approach may be warranted, as a person is ‘under so great a power, that he cannot resist or avoid’.Footnote 120 Surely Hale cannot be thinking of the impossibility of seeking official protection in extraordinary times?Footnote 121 He upholds sanctity of life in peacetime. Even then, a person under the threat of instant death would not have had recourse to official protection. Thus, Hale's bifurcated approach based on peacetime or extraordinary times is unsound, unless it is based on Montesquieu's crainte, discussed earlier. The defence of duress should not be allowed.
6. Transitional justice concerns
Are there concerns peculiar to transitional justice that weigh on the resolution of the question of whether to punish? Transitional justice seeks to signal a clean break from a predecessor regime through the restoration of justice. It aims to right the wrongs of the predecessor regime through redress for victims and prosecution of perpetrators and accomplices. I suggest that a better compromise of the conflicting concerns lies in adopting a general rule that holds individuals who reported their neighbours responsible but exempts some such individuals from punishment on limited grounds. The limited exemption could be justified by the need to demonstrate grace towards individuals who complied. For example, if a particular group of individuals had been singled out and forced to be ‘accomplices’ through an obligation to inform the authorities, in circumstances that suggest they were victimised in a different way by the predecessor despotic regime, special considerations should apply in deciding whether they should be punished as a matter of transitional justice.Footnote 122 Given the need to re-establish the value of justice, the new regime may not wish to be seen as a tormentor of a different set of victims through punishment. In the ordinary criminal context, duress has been said to be excused but not justified: the conduct is deplored but criminal responsibility is not attributed to the actor.Footnote 123 Given the pressing need to achieve transitional justice, one can draw from middle grounds in the ordinary criminal context. The actor may be held responsible for the act and conviction is used to secure a stigmatic effect, enabling the law to serve a moral signposting function, even if no, or a lesser, punishment is meted out after mitigation.Footnote 124 Alternatively, one can manage genuine hard cases by granting commutation by prerogative powerFootnote 125 or parole, exercising prosecutorial discretion, or through sentencingFootnote 126 (in cases without a mandatory sentence).
What are the mechanics of punishment, and are there special concerns that apply in the Rule X context? If Rule X was legally invalid, individuals who complied were not justified by law, but had abetted a race-motivated crime. A domestic tribunal in carrying out transitional justice may rely on the subsisting criminal law relating to abetment to murder to prosecute the individuals, or it may rely on international criminal law. Either way, the criticism may arise, as it has in the context of international criminal law, that prosecutions in such a context do not achieve intended aims. In relation to genocidaires, justifications of international criminal law, such as retribution and rehabilitation to prevent the perpetrator from reoffending, have been criticised.Footnote 127 Rehabilitation of the offender would be unnecessary in relation to the individual who complied in the Rule X context, who reported their neighbour out of fear. As for retribution, this fails to be proportionate given the many murders that have been committed by genocidaires.Footnote 128 Retribution is an ‘inherently individualizing approach to punishment’, while genocide is a systematic, mass crime with perpetrators operating within institutional or social frameworks.Footnote 129 In the Rule X context, it can be argued that categorising the complying individual's act as abetment to murder does not adequately capture the heinous essence of their complicity with the genocidaires. However, the criminal act of the ordinary individual – reporting fellow individuals – is individualised, and the punishment also is not disproportionately inadequate.
7. Conclusion
My objective has been to examine whether individuals who complied with intolerably unjust laws in a predecessor regime should be punished by a tribunal in a successor regime as a matter of transitional justice. If we regard such laws as legally invalid, the individuals’ acts of reporting their neighbours are not justified by law: they would have abetted murder or genocide and committed what is likely to be a crime under the subsisting criminal law in the predecessor regime.
The foremost jurisprudential concerns are that an individual who complied with an intolerably unjust law may not have realised it is not law and may have been acting under duress. I have examined arguments against recognising these defences in the ordinary criminal context to consider, by parity of reasoning, the appropriate approach in relation to compliance with intolerably unjust laws.
For ignorance of law, reliance on an intolerably unjust law differentiates the case from the ordinary criminal context in which individuals claim they do not know abetment to murder is illegal. However, given the egregious nature of the act commanded by the intolerably unjust law, I proposed that reliance on the purported law is unreasonable.
As for claims of duress, the less immediate nature of the threat and the greater benefits of punishing individuals who complied, when compared with the ordinary criminal context, suggest that the defence should generally be denied. Crucially, punishment as a matter of transitional justice serves a moral educative effect, sending a strong message that laws must be minimally just. Future despots in rogue states may have to commit heinous acts outside the law, as it is more likely that the legal profession, officials and individuals would resist intolerably unjust laws from the onset of a rogue regime. Exceptionally, given the objective of transitional justice, supposing those who complied were especially victimised by the predecessor regime in unusual circumstances that call for a lesser punishment, conviction may be used to secure a symbolic stigmatic effect and a lesser punishment imposed. This is so that the successor regime would not be viewed as tormenting or oppressing a different group of individuals.
I have worked with a hypothetical scenario, but only to unpack what would be at stake in punishing individuals who complied with intolerably unjust laws in similar scenarios: if the threatened sanction under the law were less severe, we would be even less sympathetic to the defence of duress being pleaded; if the injustice commanded by the purported law were less obvious, we would be more sympathetic to the claim that an individual did not know it was not law.
Acknowledgements
I appreciate the insightful comments of Professor Julian Rivers on an earlier draft of this paper, as well as those of participants of several workshops (IVR Serbia: Rule of Law Beyond the Nation-States, Edinburgh Centre of Legal Theory Seminar, and Singapore Management University School of Law Research Seminar) at which the paper was presented.
Funding statement
Not applicable.
Competing interests
The author declares none.