South Africa: Constitutional Court Support SAFLII

You are here:  SAFLII >> Databases >> South Africa: Constitutional Court >> 2000 >> [2000] ZACC 5

| Noteup | LawCite

S v Manamela and Another (Director-General of Justice Intervening) (CCT25/99) [2000] ZACC 5; 2000 (3) SA 1; 2000 (5) BCLR 491 (14 April 2000)

Download original files

PDF format

RTF format

Links to summary

PDF format

RTF format


CONSTITUTIONAL COURT OF SOUTH AFRICA




Case CCT 25/99


THE STATE

versus

SAMUEL MANAMELA First Appellant

JABULANI MDLALOSE Second Appellant

THE DIRECTOR-GENERAL OF JUSTICE Intervening Party


Heard on : 4 November 1999

Decided on : 14 April 2000




JUDGMENT






MADALA, SACHS AND YACOOB JJ:


Introduction

[1]The central issue in this matter is whether the reverse onus in a statute dealing with acquisition of stolen goods is compatible with the right to a fair trial, and, in particular, the right to silence and the presumption of innocence. The question of the constitutionality of reverse onus provisions has come to this Court with more frequency than any other matter and has produced a number of judgments in which such provisions have been struck down. The problems that they raise do not lend themselves to formulaic answers. The facts and circumstances of the present case introduce elements that compel further reflection.

[2]This case comes to us as a referral for confirmation in terms of section 172(2)(a)[1] of the Constitution, of an order made on 7 June 1999 in the Witwatersrand High Court. In that order, Tip AJ (with Wunsh J concurring) declared the reverse onus provision contained in section 37(1) of the General Law Amendment Act[2](the Act) to be unconstitutional and invalid.[3] In terms of section 172(2)(a) of the Constitution, the order has no force, unless confirmed by this Court.

[3]Section 37(1) reads as follows:

“Any person who in any manner, otherwise than at a public sale, acquires or receives into his possession from any other person stolen goods, other than stock or produce as defined in section one of the Stock Theft Act, 1959, without having reasonable cause, proof of which shall be on such first-mentioned person, for believing at the time of such acquisition or receipt that such goods are the property of the person from whom he receives them or that such person has been duly authorized by the owner thereof to deal with or to dispose of them, shall be guilty of an offence and liable on conviction to the penalties which may be imposed on a conviction of receiving stolen property knowing it to have been stolen except in so far as the imposition of any such penalty may be compulsory.” (Emphasis added.)


The relevant terms of the order are:

“2. That the words ‘proof of which shall be on such first mentioned person’ in section 37(1) of the General Law Amendment Act 62 of 1955 are declared to be inconsistent with the Constitution of the Republic of South Africa Act 108 of 1996, and accordingly invalid.

3. That in terms of section 172(1)(b) of the Constitution, the order in paragraph 2 hereof shall not invalidate any application of the reverse onus created by the words declared therein to be unconstitutional and invalid, unless:
3.1 the verdict of the trial Court was entered after 27 April 1994; and
3.2 either an appeal against or a review of that verdict is pending or the time for the noting of such appeal or the bringing of such review has not yet expired.

4. The orders set out in paragraphs 2 and 3 hereof are not of force and effect until and unless they are confirmed by the Constitutional Court pursuant to the provisions of section 172(2)(a), read with section 167(5) of the Constitution.”


The appellants seek confirmation of paragraphs 2 and 3 of the order, while the state and the intervening party (the Director-General of the Department of Justice) oppose confirmation.

[4]The crux of the issue which falls to be determined is whether the reverse onus provision contained in section 37(1) is consistent with the constitutionally entrenched right to a fair trial and, in particular, section 35(3)(h) of the Constitution which guarantees the right “to be presumed innocent, to remain silent, and not to testify during the proceedings.”

[5]For the sake of convenience, we shall refer to the parties as they were referred to in the High Court. The appellants were charged in the regional court with theft. They were arrested on 13 May 1997 in central Johannesburg while carrying boxes containing goods proved to have been stolen shortly before from a vehicle parked in the vicinity. From the record it appears that two other persons, who were walking some way ahead of the appellants, were carrying similar boxes. They managed however to avoid arrest. The appellants’ version was that one of these two other persons, known to them only as “Shorty”, a hawker, who ran a stall in Soweto where they lived, had asked them to carry the boxes for him to a nearby taxi rank. They claimed that they had neither stolen the goods nor had they known that the goods had been stolen. They had simply assumed that the goods belonged to Shorty. The magistrate accepted that their evidence in this regard could reasonably possibly be true and found them not guilty on the charge of theft. He held nonetheless that they had had physical possession of the stolen goods, which brought the provisions of section 37 into play. He rejected the evidence adduced by the appellants in support of their belief that the goods had not been stolen, and convicted them on a competent verdict of having contravened section 37.[4] Neither of the appellants was a first offender and they were sentenced to seven and six years imprisonment respectively.

[6]On 7 October 1998 they appeared unrepresented, as had been the case at their trial in the regional court, at the hearing of the appeal. During the hearing, the court itself raised the issue of the constitutionality of the reverse onus in section 37(1) which had been material to the conviction of the appellants. The hearing on appeal was postponed for the purpose of appointing counsel to represent them. At the request of the court, the Legal Resources Centre (LRC) undertook that task. Mr Hulley, instructed by the LRC, submitted heads of argument. The hearing resumed on 3 November 1998 and judgment was reserved. The High Court subsequently ordered the release of the appellants and undertook to furnish reasons at a later date. While the High Court was formulating its reasons, it came to a provisional conclusion that the reverse onus provision in section 37(1) was inconsistent with the Constitution. It invited the Department of Justice to file argument, advising it and all parties concerned that its prima facie view was that the reverse onus provision was unconstitutional. After the filing of further submissions and service thereof on all parties concerned, the matter was again heard on 20 April 1999.

[7]In its judgment, the High Court set aside the convictions and sentences on two grounds. First, a procedural exception and a finding that the appellants had not enjoyed a “just and fair trial”. The appellants had not been represented in the regional court and had not been duly warned by the magistrate of the operation of section 37(1). Second, on the ground that the reverse onus provision was inconsistent with the Constitution and as it had been decisive in the finding by the magistrate, the conviction could not be sustained.

[8]Mr D’Oliviera, on behalf of the state, and Mr Patel, on behalf of the intervening party, called upon this Court to decide two preliminary issues. First, it was argued that this matter is not ripe for hearing and second, it was contended that the constitutional issue in question is moot. It was submitted that, if successful, these preliminary arguments would dispose of the matter and obviate the need for this Court to consider the constitutionality of section 37(1).


Ripeness and Mootness

[9]Mr D’Oliviera submitted that the matter was not ripe for hearing, since this Court could not properly determine the issues before it in the absence of an analysis of the Stock Theft Act.[5] No serious argument was addressed to us in this regard and, in our view, the argument on ripeness has no merit.

[10]Mr D’Oliviera and Mr Patel proceeded to argue that the question of the validity of section 37(1) had become moot as the High Court had upheld the appeal on the basis of procedural shortcomings which violated the right of the accused to a fair trial. Counsel moreover contended that there was no pressing public need for the High Court to have considered the constitutionality of section 37(1). It was accordingly argued that this Court should not entertain the confirmation proceedings. In support of this argument, counsel referred us to various pronouncements of this Court, including Kriegler J’s statement in S v Dlamini that “this Court has long since held that, as a matter of judicial policy, constitutional issues are generally to be considered only if and when it is necessary to do so”.[6]

[11]In our view however, the issue of the constitutionality of the reverse onus was not moot in the proceedings before the High Court. That court was of the opinion that the acquittal of the appellants, on the basis of a violation of their right to a fair trial, was not dispositive of the case since the appellants could be prosecuted afresh. It accordingly went on to deal with the question of the validity of the section 37(1) reverse onus and found on the facts that the accused had been “in possession” of stolen goods. The court was, however, unable to come to a finding as to whether or not the appellants had discharged the onus of proving reasonable cause, presumably as a direct consequence of the fact that the unrepresented accused were unaware of what they had to establish to secure their acquittal, so that there was virtually no evidence on record as to why they could reasonably have believed that the goods had not been stolen. The High Court nonetheless made the following observations: first, whether or not the accused were entitled to an acquittal depended ultimately on the incidence of the onus of proof. Second, the absence of the reverse onus would lead to their acquittal since the established facts did not support a conviction if the test was the conventional one of proof beyond reasonable doubt, with the onus on the state. Ultimately then the accused would be acquitted on the merits if the state bore the conventional onus. In our view therefore, the determination of the validity of the reverse onus in section 37(1) was essential to an acquittal on the merits in the case before the High Court. It follows that the contention that the matter had become moot has no basis and must fail.

[12]In our view, there is a compelling public interest that the constitutionality of section 37(1) be determined. Continuing uncertainty in this regard may well prejudice the general administration of justice as well as the interests of the accused persons affected. As Langa J stated in S v Mbatha;S v Prinsloo:

“. . . it would be undesirable for the courts to continue applying a provision which is not only manifestly unconstitutional, but which also results in grave consequences for potentially innocent persons in view of the serious penalties prescribed.”[7]


Mr D’Oliviera advised us that there are any number of trials, either pending or proceeding, in which the reverse onus provision is liable to be invoked, so that it is necessary that legal certainty be achieved as soon as possible. In S v Hoosen,[8] Niles-Dunér J (with Pillay J concurring) held that the reverse onus provision contained in section 37(1) was an impermissible infringement of the guarantee contained in section 25(3)(c) of the interim Constitution.[9] She stated that:

“the burden upon the State of proving mens rea in the present circumstances is not so onerous or difficult to discharge that the State requires assistance by way of legislative intervention.”[10]


The conviction was set aside on that basis. Under the interim Constitution however, that court would not have had jurisdiction to make such a finding in the absence of agreement[11] of the parties. It is not clear from the judgment whether there was such an agreement.

[13]The question whether the reverse onus provision in section 37(1) infringes section 35(3)(h) of the Constitution must therefore be determined. Before doing so however, it is useful for a proper understanding of the issues to outline the history and effect of the impugned provision.


History and Effect of Section 37(1)

[14]Section 37 establishes the statutory offence of being found in possession of stolen goods. It had its genesis in pre-Union legislation which applied in the various provinces of South Africa[12] where the legislation sought to employ evidential devices which would alleviate the prosecution’s burden of proof in the common law crimes of theft and of receiving stolen goods. This was achieved by placing the burden of proof on an accused person as to the circumstances in which stolen goods were received or acquired, and the bona fide nature of his or her belief at the time of acquisition or receipt, that the goods had not been stolen. Similar legislation existed in British Bechuanaland (Botswana).[13]

[15]Section 37(1) was also enacted in response to difficulties experienced by the state in proving two of the elements of the common law offence of receiving.[14] Section 37 operates in tandem with section 36 of the Act to alleviate the usual burden of proof resting on the state in criminal trials. In circumstances where the state is unable to prove that the goods in question had been stolen, section 36 requires the prosecution simply to prove that there was a reasonable suspicion that the goods had been stolen, and that the accused was unable to give a satisfactory account of his or her possession of the goods.[15] Conversely, in circumstances where the state can prove that the goods had been stolen, but is unable to prove that the accused had known that they were stolen, section 37 places the onus of disproving the requisite mens rea on the accused.[16]

[16]Section 37(1) requires the prosecution to establish the following three elements of the offence beyond a reasonable doubt: that the accused was found in possession of goods, other than stock or produce, which were acquired otherwise than at a public sale, and that the goods had been stolen.

[17]The first and second of these elements require some elaboration. The phrase “receives or acquires into his possession” bears the same meaning in this provision as it does in relation to the common law offence of receiving.[17] In S v Zungu, Henning J held that this requirement meant that:

“. . . in order to obtain a conviction in terms of sec. 37 . . . it is incumbent upon the State to prove that the accused received stolen property into his possession in the sense that he had detentio thereof as well as physical control, where there is no proof, as in this case, of acquisition of the property. “Physical control” has not been defined in any of the cases referred to, but it seems clear that for physical control to exist it is not essential that the receiver should intend to, or in fact exercise control on his own behalf or on behalf of someone other than the person who delivered the property. If the receiver has detentio of stolen property in circumstances where he has either expressly or by necessary implication undertaken to safeguard it on behalf of the person from whom it was received, it appears to me that he has, in fact, physical control of it. The measure of such control may be limited, but once it exists, there is a receiving into possession within the contemplation of sec. 37(1) of Act 62 of 1955. Indeed, I venture to suggest that, although exceptions are conceivable, detentio would generally be accompanied by physical control.”[18]


Although different views had been expressed on the meaning of possession as contemplated by the section,[19] it is now clear from the Appellate Division decision in S v Moller[20] that the approach adopted in the Zungu case is the correct one, so that detentio coupled with physical possession will establish possession for the purposes of section 37(1). Such an interpretation has widened the ambit of section 37(1) and has led some to the view that every hotel porter or cloakroom attendant who takes property on behalf of another could conceivably risk conviction.[21] These fears were dismissed by the Appellate Division in the Moller case on the basis that, in the absence of special circumstances, there would be reasonable grounds for such a person to assume that the goods handed to him or her were not stolen.[22]

[18]The prosecution must also establish that the accused acquired the stolen goods other than at a public sale. Section 37(2) of the Act defines “public sale” as follows:

“For the purposes of subsection (1) ‘public sale’ means a sale effected —
(a) at any public market; or
(b) by any shopkeeper during the hours when his shop may in terms of any law remain open for the transaction of business; or
(c) by a duly licensed auctioneer at a public auction; or
(d) in pursuance of an order of a competent court.”


The state must establish that the acquisition of goods was neither in the form of a purchase from a shopkeeper or trader in a public market, nor a sale at a public auction. This requirement narrows the scope of section 37(1), although it does not qualify the numerous situations in which persons might receive property on behalf of others.

[19]Once the prosecution has established these three elements, the accused will be required to establish the further two elements of the offence: that the accused believed, at the time of acquiring the goods, that the person from whom he or she received them was indeed the owner of such goods, or was duly authorised by the owner to dispose of them; and that the accused’s belief was reasonable.[23] The accused must discharge this onus on a balance of probabilities. Early decisions of our courts interpreted the section as much. In R v Mdhletshe, Broome JP stated:

“In the first instance, it is only necessary for the Crown to establish that the property was stolen and that it was received by the accused. The accused is thereupon liable to be convicted unless he proves that he had reasonable cause for believing that the property belonged to the person from whom he received it.”[24] (Emphasis added.)


The Appellate Division confirmed this position in two cardinal decisions, S v Ghoor[25] and S v Moller.[26] It is, therefore, insufficient for acquittal merely to raise a reasonable doubt.

[20]Section 37(1) therefore, not only places on the accused the burden of proving the requisite mens rea on a balance of probabilities, but introduces a further departure from the common law. By imposing on the accused the burden of adducing evidence establishing the reasonableness of his or her subjective belief, the impugned provision effectively introduces statutory liability for the negligent, albeit innocent, acquisition or receipt of stolen goods. As argued by Mr D’Oliviera and Mr Patel, the inclusion of reasonableness in the subsection extends its reach to those persons who would effectively form a link in the chain of the disposal of stolen goods, even though they did not know that the goods were stolen. What is “reasonable” will depend on the circumstances. In many cases, an explanation by the accused of the manner in which the goods were acquired will be sufficient to meet the burden. Our courts approach the question whether an excuse is reasonable in the context of the character and background of the accused, the nature of the goods found in his or her possession, and the manner in which they were acquired.[27]

[21]Bearing in mind the effect of section 37(1) we turn to the question whether the reverse onus provision in section 37(1) infringes the rights of an accused person to be presumed innocent, to remain silent, and not to testify during the proceedings. And if so, whether such an infringement can be said to be reasonable and justifiable in terms of section 36 of the Constitution.


Validity of Section 37(1) in the context of the rights to silence and to be presumed innocent

[22]In this Court, Mr Trengove, together with Mr Hulley, appeared on behalf of the appellants. We are indebted to them for their assistance. They argued that the imposition of a full burden of proof upon the accused infringes the right to be presumed innocent, since it creates the risk and indeed the inevitability of a conviction, despite the existence of a reasonable doubt as to the guilt of the accused. It was on this basis that Tip AJ found the phrase “proof of which shall be such first-mentioned person” unconstitutional and invalid. Counsel made no reference to the breach of the right to silence. This issue, however, was canvassed during the hearing and needs to be addressed.

[23]In our view, the challenged phrase directly implicates the right to silence as well as the presumption of innocence, and the key to the solution of the problems raised in this matter lies in unravelling the connections between them. Both are procedural rights which are central to the adversarial criminal process which was developed under the common law and subsumed into the Bill of Rights.[28] We have concluded that, viewed in its context, and balancing all the competing factors against the background of the values of the Constitution, the challenged provision can be justified only to the extent that it infringes the right to silence. Once the objective of the statute can be met by limiting the right to silence, the core reason for breaching the presumption of innocence falls away. Our reasons for this approach and conclusion are set out below.

[24]The right to silence, seen broadly as an aspect of the adversarial trial, is clearly infringed. The inevitable effect of the challenged phrase is that the accused is obliged to produce evidence of reasonable cause to avoid conviction even if the prosecution leads no evidence regarding reasonable cause. Moreover, the absence of evidence produced by the accused of reasonable cause in such circumstances would result not in the mere possibility of an inference of absence of reasonable cause, but in the inevitability of such a finding. In these circumstances, for the accused to remain silent is not simply to make a hard choice which increases the risk of an inference of culpability.[29] It is to surrender to the prosecution’s case and provoke the certainty of conviction.

[25]Similarly the presumption of innocence is manifestly transgressed.[30] This Court has frequently held that reverse onuses of this kind impose a full legal burden of proof on the accused.[31] Accordingly, if after hearing all the evidence, the court is of two minds as to where the truth lies, the constitutional presumption of innocence is replaced by a statutory presumption of guilt. By virtue of the same logic, a conviction must follow if the court concludes that the accused’s version, even though improbable, might reasonably be true.

[26]The purpose of the presumption of innocence is to minimise the risk that innocent persons may be convicted and imprisoned. It does so by imposing on the prosecution the burden of proving the essential elements of the offence charged beyond a reasonable doubt,[32] thereby reducing to an acceptable level the risk of error in a court’s overall assessment of evidence tendered in the course of a trial. The reverse onus provision relieves the prosecution of the burden of proving all the elements of the section 37 offence by effectively presuming that any person, proven by the state to be in possession of stolen property, acquired otherwise than at a public sale, did not have reasonable cause for believing at the time of acquisition or receipt that the goods had not been stolen. Where the accused is unable to persuade the court on a balance of probabilities that reasonable cause exists, which would be the case even where the probabilities are evenly balanced, he or she must be found guilty, despite a reasonable doubt in the mind of the judicial officer as to whether or not the accused is innocent.[33] The presumption of innocence is manifestly infringed by section 37(1).[34] Unless saved as a permissible limitation, it is unconstitutional and invalid.


Justification

[27]It was argued by Mr Trengove, as a precursor to the limitation inquiry, that a statutory provision which introduces the certainty that innocent persons will be convicted is ethically offensive and can never be justified.[35] However, it is clear from the wording of section 36(1) that no right enshrined in Chapter 2 of the Constitution is absolute. Although this Court has so far not found an impugned reverse onus provision to pass constitutional muster, it has been at pains to articulate that there are circumstances in which such measures may be justifiable.[36] The effective prosecution of crime is a societal objective of great significance which could, where appropriate, justify the infringement of fundamental rights.

[28]This Court has expressly kept open the possibility of reverse onus provisions being justifiable in certain circumstances.[37] Ordinarily, a reverse onus could be justifiable only if the risk and consequences of erroneous conviction produced by a statutory presumption against the accused are outweighed by the risk and consequences of guilty persons escaping conviction simply because of categorical adherence to an impervious presumption of innocence.[38]

[29]A broad context in which the use of reverse onus provisions might be justified concerns “regulatory offences”, as opposed to “pure criminal offences”. Thus, regulatory statutes dealing with licensed activity in the public domain, the handling of hazardous products, or the supervision of dangerous activities, frequently impose duties on responsible persons, and then require them to prove that they have fulfilled their responsibilities.[39] The objective of such laws is to put pressure on the persons responsible to take pre-emptive action to prevent harm to the public.[40] Although censure might be acute, there is generally not the same stigma or the severe penalties as for common law offences. Similarly, there are cases involving the existence or authenticity of public documents or licences,[41] where practicalities and common sense dictate that, bearing in mind the reduced risk of error involved, it would be disproportionately onerous for the state to be obliged to discharge its normal burden in order to secure a conviction. Traffic regulation provides a further example,[42] such as when a statute states that the owner of a car is presumed to be the person who parked it illegally; in the great majority of cases, there is simply no way in which the state could prove who parked the car.

[30]There is also an important area in which the common law imposes an onus on the accused, namely, in relation to proof of insanity. The long-standing practice has been to require the defence to prove insanity on a balance of probabilities.[43] Here the consequences of an erroneous finding of guilt can be severe, for instance, a finding of guilt on a charge of murder or other serious crime. Balanced against this is the ease with which, it is contended, an accused could otherwise escape conviction by feigning insanity.

[31]It is not necessary in this case to decide whether the onus in each of the above situations would survive scrutiny under section 36. It is clear however that open and democratic societies permit the shifting of the burden of proof to the accused when it would not be disproportionately invasive of the right to silence and the presumption of innocence to do so. We now consider whether the section 37(1) reverse onus is justifiable.

[32]Section 36(1) of the Constitution provides:

“The rights in the Bill of Rights may be limited only in terms of law of general application to the extent that the limitation is reasonable and justifiable in an open and democratic society based on human dignity, equality and freedom, taking into account all relevant factors, including —

(a) the nature of the right;
(b) the importance of the purpose of the limitation;
(c) the nature and extent of the limitation;
(d) the relation between the limitation and its purpose; and
(e) less restrictive means to achieve the purpose.”

It should be noted that the five factors expressly itemised in section 36 are not presented as an exhaustive list. They are included in the section as key factors that have to be considered in an overall assessment as to whether or not the limitation is reasonable and justifiable in an open and democratic society. In essence, the Court must engage in a balancing exercise and arrive at a global judgment on proportionality and not adhere mechanically to a sequential check-list.[44] As a general rule, the more serious the impact of the measure on the right, the more persuasive or compelling the justification must be. Ultimately, the question is one of degree to be assessed in the concrete legislative and social setting of the measure, paying due regard to the means which are realistically available in our country at this stage, but without losing sight of the ultimate values to be protected.

[33]Although section 36(1) differs in various respects from section 33 of the interim Constitution, its application continues to involve the weighing up of competing values on a case-by-case basis to reach an assessment founded on proportionality.[45] Each particular infringement of a right has different implications in an open and democratic society based on dignity, equality and freedom. There can accordingly be no absolute standard for determining reasonableness. This is inherent in the requirement of proportionality, which calls for the balancing of different interests. The proportionality of a limitation must be assessed in the context of its legislative and social setting.[46] Accordingly, the factors mentioned in section 36(1) are not exhaustive. They are key considerations, to be used in conjunction with any other relevant factors, in the overall determination whether or not the limitation of a right is justifiable.

[34]These themes are eloquently dealt with in the judgment of O’Regan J and Cameron AJ (the minority judgment). We agree with their approach and also agree that there is a pressing social need for legislation to address the evil they identify. Section 36, however, does not permit a sledgehammer to be used to crack a nut. Nor does it allow for means that are legitimate for one purpose to be used for another purpose where their employment would not be legitimate.[47] The duty of a court is to decide whether or not the legislature has overreached itself in responding, as it must, to matters of great social concern. As the minority judgment points out, when giving appropriate effect to the factor of “less restrictive means”, the court must not limit the range of legitimate legislative choice in a specific area. The minority judgment also states that such legislative choice is influenced by considerations of cost, implementation, priorities of social demands, and the need to reconcile conflicting interests. These are manifestly sensible considerations that do not provoke disagreement. Our difference with the minority judgment is not over how the principles should be articulated, but rather as to how they should be applied in the circumstances of this case.

[35]We deal first with the right to silence. This Court has said that “[t]he right to silence, like the presumption of innocence, is firmly rooted in both our common law and statute”[48] and is “inextricably linked to the right against self-incrimination and the principle of non-compellability of an accused person as a witness at his or her trial”.[49]

[36] Pitted against this time-honoured right is the consideration that dealing in stolen property is a scourge in our society. The practice involves massive corruption and immorality that can permeate and perversely normalise itself in every area of society. The public perception that stolen goods are easily disposed of in our country, insidiously encourages serious and often violent crimes including car-jacking, mugging, robbery and theft.[50] These are indeed weighty considerations. Yet as this Court pointed out in Dlamini:
“Although the level of criminal activity is clearly a relevant and important factor in the limitations exercise undertaken in respect of [section] 36, it is not the only factor relevant to that exercise. One must be careful to ensure that the alarming level of crime is not used to justify extensive and inappropriate invasions of individual rights. It is well established that [section] 36 requires a court to counterpoise the purpose, effects and importance of the infringing legislation on the one hand against the nature and importance of the right limited on the other.”[51]

[37]The prevalence of serious crime calls for government action, but does not provide a blank cheque for the legislature to erase all procedural safeguards. Indeed, it is precisely when public emotion is at its highest that procedural protection against possible miscarriage of justice is most necessary[52]. Something case-specific and contextualised is required to bring the scales down on the side of limitation. In the present case, there are convincing reasons for an incursion into the right to silence, but not for a reverse onus which would unduly increase the risk of innocent persons being convicted.

[38]Mr D’Oliveira argued persuasively that in the vast majority of cases the state has no information or evidence concerning the circumstances in which, and the persons from whom, the accused acquired the goods in question. Almost always all the information relevant to the determination of reasonable cause is peculiarly within the knowledge of the accused. This makes it extremely difficult for the state to demonstrate the absence of reasonable cause unless there is evidence emanating from the accused. The appellants correctly did not dispute this. In the circumstances, there is nothing unreasonable, oppressive or unduly intrusive in asking an accused who has already been shown to be in possession of stolen goods, acquired otherwise than at a public sale, to produce the requisite evidence, namely, that he or she had reasonable cause for believing that the goods were acquired from the owner or from some other person who had the authority of the owner to dispose of them. For these reasons, then, the limitation on the right to silence contained in the challenged phrase is justified.

[39]The question that remains, however, is the standard by which that evidence must be evaluated. The reverse onus requires the accused to establish on a balance of probabilities reasonable cause for the requisite belief. As has been pointed out, this means that the court is obliged to convict even if it entertains a reasonable doubt as to the guilt of the accused. Indeed, the reverse onus goes further. It obliges the court to convict if the version of the accused is as likely to be as true as not. We have to decide, then, whether this limitation of the presumption of innocence can be justified.

[40]The presumption of innocence protects the fundamental liberty and human dignity of every person accused of criminal conduct. It ensures that until the state proves an accused’s guilt beyond a reasonable doubt, he or she cannot be convicted. The right is vital to an open and democratic society committed to fairness and social justice.[53] Where a presumption of guilt is substituted for the presumption of innocence, the limitation of the right is extensive and “the justification for doing so must be established clearly and convincingly”.[54]

[41]The primary ground for the justification of the section 37(1) reverse onus is the ongoing legislative endeavour to put in place effective means to eradicate the market in stolen property which has a devastating effect on the maintenance of law and order. The crimes targeted range from highly organised syndicate- and gang-related activities through corruption at various levels to those committed by people tempted by a windfall who participate, sometimes unwittingly, in the chain of disposal of stolen goods.

[42]There is little doubt that the effective prosecution of thieves and receivers dealing in stolen goods is a pressing social need. We equally accept that there are important reasons of public policy for a statutory offence that penalises those who are not dealers, but who form a link in the chain of disposal of stolen goods by deliberately or negligently failing to establish the provenance of goods they acquire outside of ordinary commercial channels. Nonetheless, the level of crime does not on its own justify any infringement of the Bill of Rights, no matter how invasive.

[43]Indeed, there are a variety of factors that point away from the conclusion that a reverse onus in this case is reasonable and justifiable. The relation between the reverse onus and the state purpose is not proportionate. Although it has been accepted that the efficacy of the section 37 offence requires it to target a wide range of people exercising physical control over stolen goods, regardless of the fact that their connection with the goods might be remote, the impugned provision is nonetheless too broadly formulated. A reverse onus proposing to criminalise the deliberate or negligent dealing in stolen goods would be easier to justify if the goods in question were confined to motor cars or expensive equipment, where members of the public could reasonably be required to document provenance, ownership and transfer. The purpose of the reverse onus in such cases would be to oblige purchasers, on pain of conviction, to keep and produce records. Yet the section 37 reverse onus applies to all kinds of goods acquired in a multitude of different circumstances and is not exclusively attracted by people and transactions where its use could most easily be justified.

[44]Because of this and also because of the manner in which “possession” has been interpreted by our courts,[55] the section 37 net extends to a wide range of people, many of whom are poor, unskilled and illiterate. It includes persons acquiring ordinary household necessities, such as clothing, food, cooking utensils and other goods from door-to-door vendors. The practical implications of this cannot be ignored. Many of these people are not likely to keep records of the wide variety of informal transactions that they conclude daily. They, and not the professional receivers, are the persons least in a position to discharge the onus and hence become the class most vulnerable to erroneous conviction precisely because of their disadvantaged position in society. Furthermore, because of their inability to afford legal representation, they will not be well prepared either to present their case to best advantage or to meet the cross-examination to come. The risk of people being erroneously convicted, subjected to the social stigma of conviction and unjustly sent to jail, is unacceptably high.

[45]It is true that judicial officers can be expected to appreciate South Africa’s social reality when applying the section. This could curtail any promiscuous recourse to the reverse onus. Similarly, temporary custodians may have little difficulty in discharging the onus. Furthermore, when passing sentence, judicial officers should distinguish between those who know full well the implications of their action, those who deliberately turn a blind eye, and those who simply do not bother to make sufficient inquiries.

[46]The presumption, however, does not entail only an assessment of whether a particular belief was reasonable in all the circumstances. Rather, and of greater significance, it obliges the accused to establish the circumstances or the factual foundation from which a conclusion as to reasonableness or otherwise may be drawn. It accordingly covers matters where the case turns substantially on the credibility of evidence given at the trial. If the accused fails to establish on a balance of probabilities the circumstances in which he or she claims to have acquired the goods (the factual foundation) a conviction will follow without the need for any further inquiry into reasonableness. That inquiry would arise only if the relevant factual foundation had first been established in favour of the accused on a balance of probabilities. Everything therefore would depend in the first place on such a credibility finding. It must be borne in mind that the probability of the accused’s version being true will be determined after cross-examination by the state which, if reasonably effected, could leave the judicial officer in a state of uncertainty as to where the truth lay.

[47]The present case illustrates the kinds of dilemmas which would face a judicial officer. It appears from rather cryptic findings that the magistrate relied heavily on the fact that the accused had failed to discharge the onus on them on a balance of probabilities. Because of the presumption, therefore, the magistrate found the accused guilty as charged and sentenced them to seven and six years’ imprisonment respectively. The situation that faced the magistrate was no different from that confronting the trier of fact in any ordinary criminal trial - was the accused telling the truth or not? There was nothing specific either in the nature of the goods or in the circumstances in which they were acquired that took the matter out of the normal forensic situation where a finding of credibility had to be made. Indeed, a reverse onus might well be easier to justify when it relates to the inference to be drawn from certain established facts than when it forces the judicial officer effectively to reject the accused’s version even if it was as probable as not. This intrudes severely on the balance between the prosecution and the accused implicit in what our Constitution regards as a fair trial.

[48]The risk that innocent persons might be convicted is aggravated by the fact that section 37 permits sentences of long terms of imprisonment and it is not open to the trier of fact to attenuate the sentences on the basis that he or she might still have had a doubt as to the guilt of the accused. Even more disturbing, the judicial officer would have to convict and impose such sentence where the accused’s story was just as likely to be true as not.

[49]In assessing whether the section 37(1) limitation of the right to be presumed innocent is reasonable and justifiable the state in this case has established the importance of the objectives sought to be attained by the impugned provision. Nonetheless, considering that the grounds of justification must be more persuasive where the infringement of the rights in question is extensive, the state has failed, in our view, to discharge the onus of establishing that the extent of the limitation is reasonable and justifiable and that the relation between the limitation and its purpose is proportional. It equally failed to establish that no less restrictive means were available to Parliament in order to achieve the purpose. The imposition of an evidential burden on the accused would equally serve to furnish the prosecution with details of the transaction at the time of acquisition or receipt. Accordingly, there is a less invasive means of achieving the legislative purpose which serves to a significant degree to reconcile the conflicting interests present in this case and which does not raise concerns relating to additional cost, the prioritisation of social demands and practical implementation.[56]

[50]In the light of the vital importance to our criminal justice system of the right to be presumed innocent and the cluster of fair trial rights which accompany it, the imposition of a full burden of proof in the circumstances has a disproportionate impact on the right in question. Had the reverse onus been wrought in a more focussed and nuanced way so as to eliminate or diminish these concerns, it might have passed scrutiny. Yet as it stands, its sweep is too great. The risk of people being erroneously convicted and unjustly sent to jail is too high. We acknowledge that ours is an open and democratic society facing many challenges with limited means, and that it is in this setting that the question of proportionality must be determined. Yet, the very circumstances that have made the challenge so great and left us with means so stretched, place those least capable of defending their rights in the greatest jeopardy of being victims of miscarriages of justice. We therefore cannot agree with the view expressed in the minority judgment that the limitation on the presumption of innocence is sufficiently focussed to be justifiable.

[51]In the present matter, we conclude that the main problem facing the prosecution will be met by requiring the accused to furnish evidence as to the reasonableness of her or his belief. We do not see any persuasive or compelling reason for reversing the usual onus of proof as well. The challenged phrase accordingly passes the test of section 36 insofar as it limits the right to silence, but not to the extent that it limits the presumption of innocence.


The Appropriate Remedy

[52]In a series of cases decided under the interim Constitution, this Court held that reverse onus provisions were inconsistent with that Constitution, that they could not ordinarily be read down to be evidential presumptions, and that they had accordingly to be declared to be invalid and of no force and effect.[57] This is the first case concerning a reverse onus provision to come before this Court since the Gay and Lesbian Immigration case[58] in which it was held that courts have the power to read words into a statute to remedy the unconstitutionality of a provision under the 1996 Constitution.

[53]The interim Constitution provided that a person whose rights under the Bill of Rights had been infringed or threatened could “apply to a competent court of law for appropriate relief, which may include a declaration of rights”. Sections 35(2)[59], 232(3)[60] and 98(5)(6)(7)[61] of the interim Constitution provided specifically for the types of orders that courts could make and for the reading down of a statutory provision in specified circumstances. It was in the light of these provisions that O’Regan J declined to read down the relevant provision and declared it to be invalid in Bhulwana.[62] That decision was followed by this Court in all subsequent reverse onus cases.[63] The appropriateness of conjoining the remedy of striking out the offending portions of a statute with that of reading words into the provision in order to ameliorate the consequences of such striking out is not dealt with in these judgements.

[54]The 1996 Constitution is different. It also makes provision for a competent court to “grant appropriate relief, including a declaration of rights” to anyone whose rights under the Bill of Rights have been infringed or threatened. It does not, however, contain specific provisions equivalent to sections 98(5), (6) and (7) which dealt with the Court’s powers in cases where laws or executive or administrative acts or conduct were found to be inconsistent with the interim Constitution. Nor are there provisions in the 1996 Constitution similar to sections 35(2) and 235(3) of the interim Constitution. Instead section 172(1) of the 1996 Constitution now affords the courts greater flexibility. It empowers and obliges a court in broad terms to make any order that is “just and equitable” if any law or conduct is declared inconsistent with the Constitution. The section provides:

“Powers of courts in constitutional matters
172(1) When deciding a constitutional matter within its power, a court-
(a) must declare that any law or conduct that is inconsistent with the Constitution is invalid to the extent of its inconsistency; and
(b) may make any order that is just and equitable, including-

(i) an order limiting the retrospective effect of the declaration of invalidity; and

(ii) an order suspending the declaration of invalidity for any period and on any conditions, to allow the competent authority to correct the defect.”

[55]In the Gay and Lesbian Immigration case[64] Ackermann J held that

“The Court’s obligation to provide appropriate relief must be read together with section 172(1)(b) which requires the Court to make an order which is just and equitable.”[65]


He went on to hold that, depending on the circumstances, reading in could be an appropriate form of relief, and that

“The real question is whether, in the circumstances of the present matter, reading in would be just and equitable and an appropriate remedy.”[66]


The reference to “just and equitable” in this passage is clearly a reference to the court’s powers under section 172(1)(b) to which he had previously referred.

[56] The principles applicable to “reading in” as a remedy for unconstitutionality are set out in the Gay and Lesbian Immigration judgment. They are:

“[74] The severance of words from a statutory provision and reading words into the provision are closely related remedial powers of the Court. In deciding whether words should be severed from a provision or whether words should be read into one, a court pays careful attention first, to the need to ensure that the provision which results from severance or reading words into a statute is consistent with the Constitution and its fundamental values and secondly, that the result achieved would interfere with the laws adopted by the legislature as little as possible. In our society where the statute books still contain many provisions enacted by a Parliament not concerned with the protection of human rights, the first consideration will in those cases often weigh more heavily than the second.

[75] In deciding to read words into a statute, a court should also bear in mind that it will not be appropriate to read words in, unless in so doing a court can define with sufficient precision how the statute ought to be extended in order to comply with the Constitution. Moreover, when reading-in (as when severing) a court should endeavour to be as faithful as possible to the legislative scheme within the constraints of the Constitution. Even where the remedy of reading- in is otherwise justified, it ought not to be granted where it would result in an unsupportable budgetary intrusion. In determining the scope of the budgetary intrusion, it will be necessary to consider the relative size of the group which the reading in would add to the group already enjoying the benefits. Where reading-in would, by expanding the group of persons protected, sustain a policy of long standing or one that is constitutionally encouraged, it should be preferred to one removing the protection completely.

[76] It should also be borne in mind that whether the remedy a court grants is one striking down, wholly or in part; or reading into or extending the text, its choice is not final. Legislatures are able, within constitutional limits, to amend the remedy, whether by re-enacting equal benefits, further extending benefits, reducing them, amending them, ‘fine-tuning’ them or abolishing them. Thus they can exercise final control over the nature and extent of the benefits.”

[57]We would add that reading in is not necessarily confined to cases in which it is necessary to remedy a provision that is under-inclusive. There is no reason in principle why it should not also be used as part of the process of narrowing the reach of a provision that is unduly invasive of a protected right. Reading down, reading in, severance and notional severance are all tools that can be used either by themselves or in conjunction with striking out words in a statute for the purpose of bringing an unconstitutional provision into conformity with the Constitution, and doing so carefully, sensitively and in a manner that interferes with the legislative scheme as little as possible and only to the extent that is essential. There is no single formula. In appropriate cases it may be necessary to delete words from a provision and read in other words to make the provision consistent with the Constitution, where the deletion of the words alone would result in the declaration of invalidity to an extent greater than that required by the Constitution. The considerations referred to in the Gay and Lesbian Immigration case would then have to be borne in mind. But if they are met there is no reason why this should not be done.

[58]The striking down of the reverse onus in section 37, without more, would leave a vacuum in the present legislative structure designed to deal with “fencing” which is a pervasive evil in our society. Parliament could remedy the situation, but that takes time, and in the interim that gap would remain. To read in the words necessary to establish an evidential presumption is less invasive of the legislative purpose of section 37 than simply striking down the presumption, and goes as far as is permissible in the context of section 37 to place a burden on possessors of stolen property to account for their possession. We therefore consider that an order in such terms would be competent and appropriate.

[59]In the result the following order is substituted for the order made by the High Court:

1. The phrase “proof of which shall be on such first-mentioned person” in section 37(1) of the General Law Amendment Act, 62 of 1955, is declared to be inconsistent with the Constitution and invalid.

1.

2. Section 37(1) should be read so as to have as a last sentence: “In the absence of evidence to the contrary which raises a reasonable doubt, proof of such possession shall be sufficient evidence of the absence of reasonable cause”.
3. The orders in paragraphs 1 and 2 above shall not invalidate any application of the reverse onus created by the words declared therein to be unconstitutional and invalid unless:
3.1 The verdict of the trial court was entered after 27 April 1994; and
3.2 Either an appeal against or review of that verdict is pending or the time for noting of such appeal or review has not yet expired.



Chaskalson P, Langa DP, Ackermann J, Mokgoro J and Ngcobo J concur in the judgment of Madala, Sachs and Yacoob JJ.

O’REGAN J AND CAMERON AJ:

[60]This case is about the circumstances in which persons found in possession of stolen goods acquired otherwise than at a public sale should be held criminally liable. The majority concludes that it is impermissible for the Legislature to require such persons, to avoid conviction, to prove that they had reasonable cause for believing at the time of acquiring the goods that they were not stolen. We disagree. In our view, it is constitutionally permissible that an accused found in possession of stolen goods so acquired be obliged to persuade a court that he or she had reasonable cause for believing at the time of acquisition that the goods were not stolen. The debate about the permissibility of imposing a burden of proof upon the accused should not shroud the real issue in this case, which in our view concerns the constitutional legitimacy of using a reverse onus to place obligations upon members of our society to ensure that they act vigilantly to inhibit the market in stolen goods.

[61]Where, as in our country, the market in stolen goods is extensive and the pattern of theft and robbery feeding that market is excessively violent, we consider that society has the right to oblige citizens to act vigilantly to ensure that they can prove that they have reason to believe that the goods are not stolen. This obligation has been imposed by the Legislature through the creation of a special offence which is tailored to capture the extent of culpability appropriate in these circumstances. The impact of the offence is that an accused, found in possession of stolen goods obtained otherwise than at a public sale and who is unable to establish reasonable cause for possessing such goods, is convicted, not of theft or common law receiving, but of a special statutory offence. In our view, there can be no constitutional complaint about this offence.

[62]Accordingly, we cannot agree with the majority that the reverse onus should be declared invalid. In our view, although the criminal offence established in section 37(1) of the General Law Amendment Act, 62 of 1955 not only trenches upon the right to silence, but also upon the presumption of innocence, it does so in a justifiable manner. We do not differ from the majority on how the matter should be approached in relation to the justifiability of the infringements in question. Where we differ is in what answer the approach should yield. We accept, for the reasons given by the majority, that to the extent that section 37 breaches the right to silence, it is justifiable. However, we disagree with the majority in that, in our view, the section’s infringement of the presumption of innocence is also justifiable. In this judgment, therefore, we consider only the latter issue — the justifiability of the breach of the presumption of innocence.

[63]Section 37(1), the terms of which for ease of reading we repeat, creates a statutory offence of being found in possession of stolen goods:

“Any person who in any manner, otherwise than at a public sale, acquires or receives into his possession from any other person stolen goods, other than stock or produce as defined in section one of the Stock Theft Act, 1959, without having reasonable cause, proof of which shall be on such first-mentioned person, for believing at the time of such acquisition or receipt that such goods are the property of the person from whom he receives them or that such person has been duly authorized by the owner thereof to deal with or to dispose of them, shall be guilty of an offence and liable on conviction to the penalties which may be imposed on a conviction of receiving stolen property knowing it to have been stolen except in so far as the imposition of any such penalty may be compulsory.” (emphasis added)


It is clear from the terms of this provision that the prosecution is required to establish beyond reasonable doubt the following three elements: that the accused has been found in possession of goods other than stock or produce,[1] that the goods had been stolen and that the accused acquired the goods otherwise than at a public sale. Only if the state has proved these elements does the accused attract the burden of establishing that he or she had reasonable cause for believing at the time of the acquisition that the person from whom the goods were received was the owner, or authorised by the owner, to dispose of them.[2]

[64]It is clear that section 37 creates a reverse onus in terms of which the accused is required to establish, on a balance of probabilities, that he or she had reasonable grounds for believing that the goods were not stolen. If the accused is unable to do so, he or she will be convicted. Even if he or she raises a reasonable possibility that such grounds existed, this will not suffice to avoid conviction. The provision may therefore result in a conviction despite the existence of reasonable doubt in the mind of the judicial officer as to whether the accused had such grounds. The provision therefore conflicts with the presumption of innocence entrenched in section 35(3) of the Constitution which reads:

“Every accused person has a right to a fair trial, which includes the right—
. . . .
(h) to be presumed innocent, to remain silent, and not to testify during the proceedings”.[3]

[65]Once it has been established that section 37 is in conflict with one of the rights entrenched in the Bill of Rights, the question arises whether it may be justified in terms of section 36(1) of the Constitution which provides:

“The rights in the Bill of Rights may be limited only in terms of law of general application to the extent that the limitation is reasonable and justifiable in an open and democratic society based on human dignity, equality and freedom, taking into account all relevant factors, including—
(a) the nature of the right;
(b) the importance of the purpose of the limitation;
(c) the nature and extent of the limitation;
(d) the relation between the limitation and its purpose; and
(e) less restrictive means to achieve the purpose.”


The approach in deciding whether a limitation is justifiable was set out in S v Makwanyane and Another as follows:

“The limitation of constitutional rights for a purpose that is reasonable and necessary in a democratic society involves the weighing up of competing values, and ultimately an assessment based on proportionality. This is implicit in the provisions of s 33(1). The fact that different rights have different implications for democracy and, in the case of our Constitution, for ‘an open and democratic society based on freedom and equality’, means that there is no absolute standard which can be laid down for determining reasonableness and necessity. Principles can be established, but the application of those principles to particular circumstances can only be done on a case-by-case basis. This is inherent in the requirement of proportionality, which calls for the balancing of different interests. In the balancing process the relevant considerations will include the nature of the right that is limited and its importance to an open and democratic society based on freedom and equality; the purpose for which the right is limited and the importance of that purpose to such a society; the extent of the limitation, its efficacy and, particularly where the limitation has to be necessary, whether the desired ends could reasonably be achieved through other means less damaging to the right in question.”[4] (footnote omitted)


In National Coalition for Gay and Lesbian Equality and Another v Minister of Justice and Others,[5] this Court held that the approach to limitations established by Chaskalson P in Makwanyane’s case applies to cases heard under the 1996 Constitution, notwithstanding the changed language of the limitation clause.

[66]The approach to limitation is, therefore, to determine the proportionality between the extent of the limitation of the right considering the nature and importance of the infringed right, on the one hand, and the purpose, importance and effect of the infringing provision, taking into account the availability of less restrictive means available to achieve that purpose. The limitation analysis that follows will therefore first consider the extent of the limitation of the right caused by section 37, and will then turn to the purpose, importance and effect of section 37. These are the two issues whose relative weight determines the outcome of the limitation analysis. That analysis therefore concludes by comparing the relative weight. As stated above, our difference with the majority lies not in the approach to the limitation analysis, but in its application in the current case.


The scope of the infringement and the nature of the presumption of innocence

[67]We now consider the first aspect. It is clear from the above analysis that section 37 infringes the right to be presumed innocent. The risk that an accused person will be convicted despite the existence of a reasonable doubt is clearly an infringement of the presumption of innocence which is a fundamental principle of our criminal justice system.[6]

[68]The presumption of innocence and, in particular, rules concerning the burden of proof (both its incidence and the standard required) exist because fact-finding by a court can never be without the risk of error[7] and because, at times, courts cannot determine the facts at all. Rules regulating the burden of proof seek to determine the acceptable level of risk and who should bear it in each case. Because our Constitution and criminal justice system seek to avoid the conviction of the innocent, we observe and defend the presumption of innocence which requires that guilt be established beyond reasonable doubt before the accused can be convicted. We entrench the presumption of innocence in our Constitution to remind us that we wish to minimise as much as is reasonably possible the risk of error in the proceedings that determine whether a person is to be punished by the state for criminal conduct and to ensure that an accused is reasonably protected from the risk of error. However, like other rights, the presumption of innocence is not absolute. Once it is infringed, the question arises whether there are important reasons, outweighing the importance of the presumption, for increasing the risk of error through varying the burden or incidence of proof. This Court has always acknowledged that there are circumstances in which a burden of proof may be imposed upon an accused. Whether such circumstances have arisen is a matter to be determined under section 36 of the Constitution.[8]

[69]The level of justification required to warrant a limitation upon a right depends on the extent of the limitation. The more invasive the infringement, the more powerful the justification must be.[9] It is important to recognise that not every reverse onus offends the presumption of innocence in the same manner or to the same extent. To assess the extent of the limitation it is necessary to examine carefully the legislative provision in question.

[70]The presumption in this case affects only those who have obtained stolen goods but who did not buy them at a public market or from a shopkeeper or at an auction. Section 37(2) provides as follows:

“For the purposes of sub-section (1) ‘public sale’ means a sale effected—
(a) at any public market; or
(b) by any shopkeeper during the hours when his shop may in terms of any law remain open for the transaction of business; or
(c) by a duly licensed auctioneer at a public auction; or
(d) in pursuance of an order of a competent court.”

[71]People who acquire goods in the circumstances in section 37(2) will not be required to provide cause for believing that the goods were not stolen. Those circumstances are widely defined to include all conceivable occasions where the public can without anxiety or suspicion purchase items in common commercial traffic. The state retains the burden of proving this element beyond reasonable doubt. Only once it has done this is the accused put on risk of criminal conviction in the absence of an account establishing that there was probably reasonable cause for believing that the disposer of the goods had lawful power to pass them on.

[72]On whether the accused had reasonable grounds for believing the goods not to be stolen the state will rarely be able to lead evidence in contradiction. The effect of section 37 is that the accused will need to establish first that the explanation is true and second that it is reasonable.1[0] In relation to the first issue, the accused will be convicted where the court disbelieves the version advanced, even though accepting that it may reasonably possibly be true, or where the court cannot decide whether the version is true or not. In relation to the second issue, the court will convict the accused even where it accepts that the explanation is true if it is not satisfied that the explanation is reasonable. Because the state will rarely lead evidence in contradiction, in the great majority of section 37 cases the key issue will be whether the accused is found to be credible or not. In this it is hard to envisage many cases where the difference between raising a reasonable doubt and establishing a reasonable probability will be of significant moment.

[73]The concept of “reasonableness” moreover bears upon the question whether the reverse onus in section 37 creates an unacceptable risk of convicting innocent persons. In considering the justification for the infringement the reverse onus entails, a telling consideration mentioned by the majority is that many of those who are most prone to fall within its net are poor. Two considerations in our view suggest however that concern in this regard should not be overstated.

[74]The first is that prudent application of section 37’s requirement of “reasonable cause” appreciably reduces the risk of unfair convictions. The requirement of reasonable cause introduces an objective element into the analysis. An accused is required to establish that the grounds proffered for believing the goods were not stolen would have been accepted by a reasonable person as grounds for that belief. The difficulties of applying a purely objective test in a diverse society have been acknowledged by our courts1[1] and have led some commentators to suggest that the test for culpa in our law should be subjective.1[2] Whatever the merits of this suggestion, it is clear that in applying the “objective” element in the determination of reasonable cause, the court does not ignore the material circumstances in which the accused found himself or herself.1[3] In R v Mbombela, one of the early authoritative cases establishing the objective criterion, the court held that —

“[a] reasonable belief, in my opinion is such as would be formed by a reasonable man in the circumstances in which the accused was placed in a given case”.1[4] (emphasis added)

[75]The approach in Mbombela’s case has been followed repeatedly.1[5] In S v Van As, Rumpff CJ explained the origin and application of the frequently-invoked standard of the “careful head of a family”, the diligens paterfamilias. He stated:

“In our law since time immemorial we have used the diligens paterfamilias as someone who in specified circumstances would behave in a certain way. What he would do is regarded as reasonable. We do not use the diligentissimus [excessively careful] paterfamilias, and what the diligens paterfamilias would have done in a particular case must be determined by the judicial officer to the best of his ability. This diligens paterfamilias is of course a fiction and is also, all too often, not a pater [father]. In the application of the law he is viewed ‘objectively’, but in essence he must apparently be viewed both ‘objectively’ and ‘subjectively’ because he represents a particular group or type of persons who are in the same circumstances as he is, with the same ability and knowledge. If a person therefore does not foresee what the other people in his group in fact could and would have foreseen, then that element of culpa, that is failure to foresee, is present.”1[6] (our translation)
[76]The test for reasonableness, of course, remains objective. But what is reasonable will be construed in the circumstances in which the accused in a particular case finds himself or herself. The courts will therefore take into account the circumstances in which the accused acted in determining whether it was reasonable to believe that the goods were not stolen. “Reasonableness” is a legal commonplace in the courts which are required to apply it daily in determining the standard of care exacted of persons in ordinary life. Whether on the facts established an accused had “reasonable cause” will depend upon the presiding officer exercising a sound and fair judgment in regard to a number of factors including —
• the nature and value of the goods acquired;
• how they were acquired and the price, if any, that was paid for them;
• the person from whom they were acquired;
• the manner in which trade in such goods normally occurs;
• the volume in which the goods in question are traded; and
• the social context in which the acquisition occurs.

[77]The question that divides this Court is whether the test for acquittal should be a reasonable possibility — as opposed to a probability — that the accused had reasonable cause for believing the goods were not stolen. This in our view depends less on the elucidation of sophisticated legal formulae than on the practical employment of good sense in the lower courts, where prosecutions under section 37 are generally brought. If an insufficiency of good sense there leads to imprudent convictions, this is a matter for correction in the appellate courts, which can be relied upon to establish sensible guidelines for when it can be concluded that accused persons had “reasonable cause” for believing goods they acquired were not stolen.

[78]A second important consideration in determining the justifiability of section 37’s infringement of the presumption of innocence is the seriousness of the offence it creates and the accompanying question of sentence. In this regard we differ from the majority, whose analysis in our view tends to overstate the seriousness of the offence. Section 37 does not render a convicted accused guilty of common law theft, nor even of common law receiving. That the legislature considered a contravention of this section as being less serious than either is spelt out in the provision itself, which renders an accused —

“liable on conviction to the penalties which may be imposed on a conviction of receiving stolen property knowing it to have been stolen except insofar as the imposition of any such penalty may be compulsory.” (emphasis added)

[79]Recognising this, the courts have already established a realm of negligent as opposed to dishonest contravention of section 37, and marked that out as deserving special consideration in regard to punishment. In S v Ghoor, Holmes JA held that where an accused subjectively believed that the goods were not stolen, but was unable to prove that reasonable grounds existed for this belief, the crime committed was “not a question of dishonesty, but more a matter of negligence”.1[7] The prison sentence imposed by the trial court was set aside on appeal and replaced with a fine and a suspended term of imprisonment. The basis upon which an accused is convicted is thus determinative, as in Ghoor, of the question of sentence. In the present case, the long prison sentences imposed were the result of the previous convictions of the two accused.

[80]The legal implications of section 37 are complex. A court is, of course, obliged to explain those implications fully to an unrepresented accused and a failure to do so may, as the High Court held in this case, vitiate a conviction. Section 35(3)(g) of the Constitution requires that if substantive injustice would otherwise result, an accused person is entitled to legal representation at state expense. It is the task of the judicial officer presiding at the trial to ensure that this right is not breached.

[81]All these considerations lead us to conclude that the risk of unfair convictions under section 37’s reverse onus should not be overstated. In this regard, the character and effect of the onus is somewhat different from an onus on a factual matter on which the state may have access to conflicting evidence. In some of the cases in which this Court has considered the constitutionality of provisions imposing an onus upon the accused, the character of the onus has related directly to factual matters where it could be expected of the state, if it wished to secure a conviction, to adduce evidence of its own, and where the imposition upon the accused of the burden of proof implausibly and unjustly stretched the bounds of inference. For example, in Bhulwana,1[8] once the accused was shown to have been found in possession of a specified (small) quantity of marijuana, the accused bore the onus of establishing that he or she was not dealing in that substance. The inference was not only lacking in logic, but the state could, if it wanted to convict the accused as a dealer, have been expected to produce evidence of dealing. Similarly, in Mbatha,1[9] once an accused was shown to be in the vicinity of firearms, the accused was required to establish that he or she was not in possession of the firearms. Again, the inference of possession was strained, and the state could have been expected to produce evidence of possession itself. In these cases, the onus relates to a factual matter on which the state could be expected to lead evidence contradicting the accused’s denial of dealing or possession. In the present case, not only is the manner in which the accused came into possession of the stolen goods uniquely within the accused’s knowledge, but in the absence of explanation the inference that a culpable connection exists between the accused and the criminal conduct that deprived the rightful owner of the goods will generally follow. It is against this background that the constitutional legitimacy of section 37 must be assessed.

[82]It is important to appreciate the specific character of the offence section 37 creates. In effect, the legislature has criminalised possession of stolen goods where an accused cannot establish reasonable cause for possessing them. The purpose of the offence is clear: it is to regulate the market in stolen goods by imposing obligations upon members of the public to act diligently by avoiding participation in that market. The method section 37 uses to achieve this objective is to oblige someone caught in possession of stolen goods, acquired otherwise than at a public sale, upon pain of criminal punishment to advance a reasonable and probable explanation for their possession. In doing so, the state imposes a burden on that person in the sense that a reasonably possible explanation — in other words, a reasonable possibility of having reasonable cause — will not suffice to escape criminal conviction. The explanation must also be probable. The statutory offence of which the accused is convicted is, in effect, that of being unable so to satisfy a court.

[83]The infringement of the presumption of innocence lies in the fact that, once the state has proved possession of stolen goods acquired otherwise than at a public sale, the offence requires the accused to establish probable and reasonable cause, rather than requiring the state to prove its absence. But the accused is convicted not of theft or of knowing receiving but of a lesser offence. It was not argued, nor could it have been, that the creation of such an offence gave rise to a conflict with other provisions of the Bill of Rights (such as the right to liberty).2[0] The only issue before us is whether the form of the offence, to the extent that it contains a reverse onus provision, is constitutionally repugnant. In contrast with the reverse onuses this Court rejected as unconstitutional in Bhulwana and Mbatha, the special statutory offence created seems to us neither inherently illogical nor inherently unjust.

[84]At the first step in the proportionality test we therefore conclude that although section 37 does infringe the presumption of innocence, the infringement is materially different from, and appreciably less invasive than, the infringement caused in many of the earlier cases concerning reverse onuses heard by this Court.


The Purpose and Effect of Section 37

[85] We turn now to the second stage of the limitation analysis in which we focus upon the purpose and effect of section 37 and the existence of less restrictive means.

[86]It is important in this regard to denote precisely the area in which the presumption in section 37 operates to assist the state. Where an accused is caught in possession of stolen goods, their mere possession may, by itself, give rise to an inference that the accused is criminally connected with the unlawful removal or receipt of the goods. If the accused is caught soon after the goods are stolen, common sense may lead to the conclusion that the only reasonable inference is that he or she stole them or participated in their theft.2[1] If the period between theft and apprehension is longer, in the absence of a satisfactory explanation the appropriate inference may be that the accused is guilty of the common law offence of receiving stolen property knowing it to be stolen.2[2] The closer the proximity in time between theft and possession, the more easily the state will be able to rely upon an inference of criminal conduct on the part of the accused. In all these cases, however, the conviction of theft or criminal receiving depends upon the state being able to establish the requisites of the crime beyond reasonable doubt: an inferential probability does not suffice. Where the time lapse is so great that an inference of theft or related criminal conduct or knowing receipt cannot be drawn at all, the state’s predicament is great. The accused can in these cases with relative ease advance a trumped-up story relating to the acquisition of the goods with little risk that the state will be able to rebut it to the requisite degree of proof.

[87]It is for this reason that the legislature chose to enact section 37, which, as we have sought to show, is closely tailored to meet the state’s difficulties without unduly menacing the public with criminal consequences. Mr D’Oliveira argued on behalf of the National Director of Public Prosecutions that the underlying purpose of section 37 was to eradicate the trade in stolen goods and to deter people from being willing to safeguard stolen goods on behalf of thieves or other persons. If the market for stolen goods were to be destroyed, and if people were to refuse to safeguard stolen goods, he argued, the primary crimes of robbery and theft would significantly be reduced. He relied on the following statement by Kriegler J in S v Bequinot:

“The receipt of stolen goods is a vital link in the chain of gainful disposal of the spoils of criminality. It is, of course, also a powerful incentive to such criminality and statutory devices aimed at facilitating the successful apprehension and prosecution of receivers of stolen property, such as s 37 clearly is, cannot lightly be invalidated.”2[3]

[88]We agree with Mr D’Oliveira that the primary purpose of section 37 is to discourage people from acquiring goods, otherwise than at a public sale, without first ascertaining satisfactorily that the goods have not been stolen. The brute and obvious corollary is that if the public undertakes such inquiries as the provision reasonably seeks to impose, traffic in stolen goods will diminish. Nor is it by any means fanciful to suppose that if the traffic in stolen goods is arrested, the instigating sources of that traffic — including the violent depredations through armed robbery and hijacking that have become the bane of too many South Africans’ lives — will also be inhibited. The importance of the purpose section 37 avows is therefore beyond dispute.

[89]The means chosen by section 37 is, as explained above, to require accused persons found in possession of stolen property acquired otherwise than at a public sale to persuade a court that at the time they acquired the property they had reasonable grounds for believing the goods were not stolen. The effect of the presumption is to require members of the public to take care when purchasing or acquiring goods otherwise than at a public sale to ensure that they take steps that will enable them to establish that they had reasonable cause to believe that the goods are not stolen. This is the very purpose of section 37 and it is precisely what the section achieves. It requires all of us not to turn a blind eye to the dubious origin of goods proffered for sale. Instead we must act diligently to ensure that in purchasing or acquiring goods otherwise than at a public sale we are not contributing to the market for sale of stolen goods which is a primary cause of crime nor are we assisting criminals by negligently safeguarding goods that are the proceeds of crime. The section was enacted in recognition of the fact that it is not only the state that must seek to combat crime. All law-abiding citizens must take the necessary steps to discourage criminal conduct and to refrain from implicating themselves in its ambit. The legislative purpose underlying section 37 is achieved precisely by imposing a burden of proof upon the accused.

[90]The majority concludes that the presumption imposed by section 37 is not justifiable because it captures within its ambit a wide range of innocent people as well as a wide range of stolen goods. All of the people who fall within the terms of the section are people who have been proved to have been in possession of stolen goods. The category of persons targeted by the provision is thus precisely defined. They are moreover the people best placed to explain the manner of their acquisition. There is therefore a close rational connection between those targeted by the section and the substance of the reverse onus provision.2[4] What section 37 requires accused people to establish is that when they took possession of the goods it was reasonable to believe the goods were not stolen. In S v Moller, the Appellate Division held that a porter or cloakroom attendant who takes possession of goods will, in the absence of special circumstances, act reasonably in believing that the goods are not stolen.2[5] There will be many other cases where it will be easy to establish that it was reasonable to assume that the goods were not stolen. In such cases, the account of the accused person concerning the circumstances of acquisition of the goods will be sufficient to discharge the burden that the statute imposes.

[91]In any event, there seems to be no reason why the Legislature should be required to limit the scope of the provision only to those persons who obtained the stolen goods for gain. As the purpose is to reduce the number of property offences, it is necessary to target all persons who may take the goods to keep them safe, whether knowingly or unknowingly, on behalf of the thief or thieves, and not only those who purchase the goods subsequently. It is necessary not only to punish the improvident purchase of stolen goods, but also those who improvidently safeguard such goods. We are not persuaded that it is necessarily more culpable for a person to purchase stolen goods without due circumspection than to take them into safekeeping without due circumspection. Similarly, it is our view that the section’s unlimited application to all kinds of stolen goods serves an important purpose. If the provision were to have the narrower effect proposed by the majority it would not achieve the broader purpose sought by the legislature. In our view that purpose is legitimate.

[92]The presumption in this case also requires an accused to give evidence of a matter which it would be unreasonable to expect the state to be able to prove. As Mr D’Oliveira contended, when a person is found in possession of stolen goods the state usually has no way of determining how that person came into possession of such goods. Indeed this is an example of a matter where the state is generally not in a position to produce evidence unless the accused speaks. It was in respect of precisely such circumstances that Kentridge AJ in S v Zuma reasoned as follows:

“Some [presumptions] may be justifiable as being rational in themselves, requiring an accused person to prove only facts to which he or she has easy access, and which it would be unreasonable to expect the prosecution to disprove. ”2[6]

[93]The effect of requiring the state to bear the burden may mean that the offence would only be effectively prosecuted in extremely rare circumstances, if at all. For that reason, it falls as well within a class of exception identified by Kentridge AJ in Zuma:

“[T]here may be presumptions which are necessary if certain offences are to be effectively prosecuted, and the State is able to show that for good reason it cannot be expected to produce the evidence itself. The presumption that a person who habitually consorts with prostitutes is living off the proceeds of prostitution was upheld on that basis in R v Downey . . . by the Supreme Court of Canada. A similar presumption in a United Kingdom statute was upheld by the European Court of Human Rights in X v United Kingdom”.2[7] (footnote omitted)

[94]The majority considers that the purpose and effect of section 37 could be achieved by less restrictive means and proposes accordingly to cure the provision by reading in words that would impose an evidential burden upon the accused. It is clear that the question whether there are less restrictive means to achieve the government’s purpose is an important part of the limitation analysis. However, it is as important to realise that this is only one of the considerations relevant to that analysis. It cannot be the only consideration. It will often be possible for a court to conceive of less restrictive means, as Blackmun J has tellingly observed:

“And, for me, ‘least drastic means’ is a slippery slope . . . A judge would be unimaginative indeed if he could not come up with something a little less ‘drastic’ or a little less ‘restrictive’ in almost any situation, and thereby enable himself to vote to strike legislation down.”2[8]

[95]The problem for the Court is to give meaning and effect to the factor of less restrictive means without unduly narrowing the range of policy choices available to the Legislature in a specific area. The Legislature when it chooses a particular provision does so not only with regard to constitutional rights, but also in the light of concerns relating to cost, practical implementation, the prioritisation of certain social demands and needs and the need to reconcile conflicting interests. The Constitution entrusts the task of legislation to the Legislature because it is the appropriate institution to make these difficult policy choices. When a court seeks to attribute weight to the factor of “less restrictive means” it should take care to avoid a result that annihilates the range of choice available to the Legislature. In particular, it should take care not to dictate to the Legislature unless it is satisfied that the mechanism chosen by the Legislature is incompatible with the Constitution.

[96]In our view, the question whether the purpose of a specific legislative provision can be achieved through less restrictive means requires a careful analysis of the purpose of the provision. As we have reasoned above, section 37 seeks to weaken the market for stolen goods by imposing an obligation upon members of the public. In effect, it requires them to be sure when they acquire goods in the circumstances broadly defined in section 37(2) that the provenance of the goods is sound; and it achieves this goal by the threat of criminal conviction and punishment in the absence of reasonable and probable cause to that effect. To the extent that section 37 imposes this obligation upon the public, it deters improvident acquisition of goods other than at a public sale by compelling members of the public to satisfy themselves in a manner that they consider will subsequently satisfy a court that it is reasonable to believe that the goods are not stolen. The reverse onus imposed by section 37 seeks to compel members of the public not to turn a blind eye to the possibility that goods they acquire may be stolen. In effect, section 37 imposes an obligation of due diligence upon all members of our community to take care to avoid participating in a market for stolen goods.2[9]

[97]The majority proposes that an evidential burden would be less invasive of the accused person’s rights while still achieving the purpose of section 37. We cannot agree that it will still fully achieve that purpose. Requiring the accused to bear only the evidential burden would have the effect of requiring the accused merely to raise a reasonable doubt as to whether there were reasonable grounds for believing the goods were not stolen. Unlike the reverse onus, this does not require the accused to establish that there were such reasonable grounds. In our view, this expedient is not a fully adequate alternative to the reverse onus for it falls short of achieving the important legislative purpose. That purpose seeks to impose an obligation upon members of the public, where stolen goods are acquired otherwise than at a public sale, to produce probable proof to escape criminal conviction. If the public realise that they bear this obligation, it is not unrealistic to expect that the traffic in stolen goods can be diminished. A requirement that mere reasonable doubt suffices constitutes a far weaker message to the public. An evidential burden, therefore, diminishes the obligation upon members of the public to act vigilantly to avoid furthering traffic in stolen goods, and thus thwarts the legitimate purpose of section 37.3[0]

[98]Where an accused has to prove on a balance of probabilities that it was reasonable to assume that the goods were not stolen, and not merely raise a reasonable doubt in that regard, there will be occasions where an accused may have to produce documentary or other real evidence to meet that burden. An accused found in possession of a stolen motor vehicle may have to do more than state when and where he or she purchased the vehicle. For example, it may be necessary to disclose the documentation for re-registration of the vehicle. In many situations, therefore, the presumption section 37 creates will require people to take steps to show that the belief they held was reasonable. This seems to be an entirely legitimate state objective, pursued by reasonable means.


Proportionality analysis

[99]If we then weigh the scope of the infringement of the presumption of innocence against the purpose, importance and effect of section 37, it is our view that the scale is tilted in favour of the constitutionality of section 37. The need to discourage improvident acquisition of stolen goods by imposing an obligation upon members of the public to take diligent care when acquiring goods and to satisfy themselves that reasonable grounds for believing that the goods are not stolen can later be shown, is of cardinal importance in a society like ours, racked as it is by high levels of property-related crime often accompanied by horrifying violence. We acknowledge that section 37 does infringe the presumption of innocence and does impose an obligation upon an accused to establish that he or she had reasonable grounds for believing goods not to be stolen. There can be no doubt that, as a general rule, it is inappropriate for an obligation to be placed upon an accused to establish innocence. However, it is our view that a limitation on the presumption of innocence that results in a duty of vigilance, coupled with an obligation to persuade a court that in acquiring goods one has acted responsibly, in order to achieve the overall purpose of smothering the market in stolen goods, is justifiable.

[100]In a society beset by robbery and theft, in which there is an active market for stolen goods, it is not unjustifiable for the legislature to say to citizens: take care not to encourage this market in stolen goods since it is the very existence of this market that gives rise to crime — if you are found in possession of stolen goods obtained otherwise than at a public sale, you will be guilty of a statutory offence unless you can show that you had reasonable grounds to believe at the time you obtained the goods that they were not stolen. Such an exhortation recognises that the protection of individual rights depends not only on the actions of the state, but on the actions of fellow citizens. The conduct of each individual can and will contribute to a climate in which the rights of others are respected. Our society asserts individual moral agency and it does not flinch from recognising the responsibilities that flow from it. It is upon this principle that democracy and respect for human rights are built. As Honoré has recently observed:

“[W]e do well, indeed we are impelled . . . to treat ourselves and others as responsible agents. But the argument for welcoming this conclusion is not that our behaviour is uncaused — something that we cannot know and which, if true, would be a surprise — but that to treat people as responsible promotes individual and social well-being. It does this in two ways. It helps to preserve social order by encouraging good and discouraging bad behaviour. At the same time, it makes possible a sense of personal character and identity that is valuable for its own sake.”3[1]


The employment of a criminal sanction as an adjunct to the assertion of these values as in this case is not in our view constitutionally illegitimate.

[101]As stated above, the purpose of the presumption of innocence is to minimise the risk of the conviction of persons who are not guilty. Section 37 does mean that in some cases members of our society who obtained stolen goods believing that they were of pure provenance, but could not establish that this was reasonable, will be convicted of a statutory offence. But after careful analysis of section 37 and its purpose, we are not persuaded that the risk of error it introduces is inappropriate in the light of the purpose it serves. Given the flexible mould of reasonableness in which the offence is cast, the experience which our courts have in applying the concept of reasonableness, the diminished opprobrium that a conviction of the statutory offence as opposed to the common law crimes of theft and receiving attracts and the concomitantly lessened penalties, we consider that the means employed, of reversing the onus, is legislatively well warranted.

[102]For these reasons, we conclude that section 37 is not unconstitutional and we would decline therefore to confirm the order of the High Court.

For the appellants : W Trengove SC and G Hulley instructed by the Legal Resources Centre.

For the state : JA van S d’Oliveira SC, P Louw and E Matzke instructed by the State Attorney, Pretoria.

For the intervening party : EM Patel SC and P Dhlamini instructed by the State Attorney, Johannesburg.


[1] Section 172(2)(a) provides:
“The Supreme Court of Appeal, a High Court or a court of similar status may make an order concerning the constitutional validity of an Act of Parliament, a provincial Act or any conduct of the President, but an order of constitutional invalidity has no force unless it is confirmed by the Constitutional Court.”
[2] Act 62 of 1955.
[3] This judgment is reported as S v Manamela and another 1999 (9) BCLR 994 (W).
[4] Compare sections 260(f) and 264 of the Criminal Procedure Act 51 of 1977.
[5] Act 57 of 1959.
[6] S v Dlamini; S v Dladla and others; S v Joubert; S v Schietekat [1999] ZACC 8; 1999 (4) SA 623 (CC), 1999 (7) BCLR 771 (CC) at para 27 (footnote omitted).
[7] [1996] ZACC 1; 1996 (2) SA 464 (CC), 1996 (3) BCLR 293 (CC) at para 30.
[8] 1999 (9) BCLR 987 (N).
[9] The Constitution of the Republic of South Africa, Act 200 of 1993.
[10] S v Hoosen above n 8 at 992E-F.
[11] See section 101(6) of the interim Constitution.
[12] See Stock Theft Act 57 of 1959; Stock Theft Act 26 of 1923; Ordinance 6 of 1904 - Transvaal; Ordinance 52 of 1903 - Orange River Colony; Cattle Stealing Act 1 of 1899 - Natal; Theft of Stock and Produce Act 35 of 1893 - Cape of Good Hope.
[13] Proclamation 141 BB, 15 March 1872 - British Bechuanaland.
[14] Someone who acquires or receives stolen property, knowing it to be stolen, is guilty of the common law offence of receiving. To secure a conviction, the prosecution has to prove, amongst other things, that the property in question had been stolen and that the accused knew, at the time of taking receipt of the property, that it had been stolen (Milton South African Criminal Law and Procedure 3 ed vol II Common-Law Crimes (Juta, Cape Town 1996) at 664-6; Snyman Strafreg 4 ed (Butterworths, Durban 1997) at 525-6).
[15] See S v Osman and another v Attorney-General, Transvaal 1998 (4) SA 1224 (CC), 1998 (11) BCLR 1362 (CC). Section 36 of the Act reads:
“Any person who is found in possession of any goods, other than stock or produce as defined in section one of the Stock Theft Act, 1959 (Act 57 of 1959), in regard to which there is a reasonable suspicion that they have been stolen and is unable to give a satisfactory account of such possession, shall be guilty of an offence and liable on conviction to the penalties which may be imposed on a conviction of theft.”
[16] In S v Zungu 1972 (3) SA 44 (N) at 50A, Henning J stated:
“I also associate myself with the view . . . that the Legislature seems to have intended by sec. 37 ‘not to change the common law on “receiving possession” but only on mens rea and onus’.”
[17] S v Moller [1990] ZASCA 62; 1990 (3) SA 876 (A) at 884 I-J; R v Von Elling 1945 AD 234 at 251.
[18] Above n 16 at 49E-H. Approved in S v Moller above n 17 at 887D. See also S v Moniz 1982 (1) SA 41 (C) at 47G - 48C.
[19] See, for example, S v Mtolo 1963 (3) SA 676 (T) at 679B-C; S v Nel 1963 (1) SA 383 (T) at 385A-C.
[20] Above n 17 at 887E.
[21] S v Mtolo above n 19 at 679C-D.
[22] Above n 17 at 886F-G.
[23] Some academic commentators have pointed to a lack of clarity in the case law as to whether an accused person has to show both that he or she believed that the goods were not stolen and that such belief was reasonable, or that the accused merely has to show that there were reasonable grounds for believing that the goods were not stolen. They are inclined, however, to the view that, once an accused has succeeded in establishing that there were reasonable grounds for believing that the goods were not stolen, it will not be difficult for the same accused to establish his or her subjective belief in that regard (Hiemstra Suid-Afrikaanse Strafproses 4ed (Butterworths, Durban 1987) at 824 -5; Snyman Strafreg 3 ed (Butterworths, Durban 1992) at 527, Milton above n 14 at 677-8). Little turns on this dispute for current purposes.
[24] 1957 (3) SA 291 (N) at 292A-B. See also S v Mtolo above n 19 at 678A-B.
[25] 1969 (2) SA 555 (A) at 557F: “En die las het op die appellant gerus om beide van hierdie faktore op ’n oorwig van waarskynlikhede te bewys.”
[26] Above n 17 at 888A-B:
“Die applikant was dus beswaar met ’n bewyslas om te bewys dat hy redelike gronde gehad het om ten tyde van verkryging of ontvangs aan te neem dat die goed die eiendom was van Heyns (van wie hy gesê het dat hy dit ontvang het) of dat Heyns gemagtig was om daaroor te beskik of om dit van die hand te sit.”
[27] See, for instance, the approach taken by the Appellate Division in the Moller case above n 17 at 886F-G.
[28] Section 35(3) of the Constitution. See S v Zuma and Others [1995] ZACC 1; 1995 (2) SA 642 (CC); 1995 (4) BCLR 401 (SA) at para 33 and S v Baloyi (CC) Case CCT 29/99, 3 December 1999, as yet unreported at para 15.
[29] In contrast with Dlamini above n 6 and Osman above n 15.
[30] See Zuma above n 28; S v Bhulwana; S v Gwadiso [1995] ZACC 11; 1996 (1) SA 388 (CC), 1995 (12) BCLR 1579 (CC); Mbatha above n 7; S v Julies [1996] ZACC 14; 1996 (4) SA 313 (CC); 1996 (7) BCLR 899 (CC); Scagell and others v Attorney-General, Western Cape and others [1996] ZACC 18; 1997 (2) SA 368 (CC), 1996 (11) BCLR 1446 (CC); S v Ntsele [1997] ZACC 14; 1997 (2) SACR 740 (CC), 1997 (11) BCLR 1543 (CC); S v Mello and another 1998 (3) SA 712 (CC), 1998 (7) BCLR 908 (CC).
[31] See Zuma above n 28 at para 19; Bhulwana above n 30; Julies above n 30; Mbatha above n 7.
[32] Compare the decision of the Supreme Court of Canada in R v Oakes (1986) 26 DLR (4th) 200 at 214 in which it was held that the presumption of innocence contains three fundamental components: the onus of proof lies with the prosecution; the standard of proof is beyond a reasonable doubt; and the method of proof must accord with fairness.
[33] S v Ghoor above n 25 at 557; S v Hoosen above n 8; S v Kaplin and others 1964 (4) SA 355 (T) 357-8.
[34] See Mello above n 30; S v Coetzee and others [1997] ZACC 2; 1997 (3) SA 527 (CC), 1997 (4) BCLR 437 (CC); Ntsele above n 30; Scagell above n 30 ; Julies above n 30; Mbatha above n 7; Bhulwana above n 30 at para 16; Zuma above n 28.
[35] See also Paizes “A closer look at the presumption of innocence in our Constitution: what is an accused presumed innocent of ?” (1998) 11 SA Journal of Criminal Justice 409; “Chasing shadows: exploring the meaning, function and incidence of the onus of proof in the South African law” (1999) 116 SA Law Journal 531.
[36] Zuma above n 28 at paras 19 and 41; see also S v Meaker 1998 (2) SACR 73 (W), 1998 (8) BCLR 1038 (W). As was stated by the European Court of Human Rights in Salabiaku v France [1988] ECHR 19; 13 EHRR 379 at 388: “Presumptions of fact or of law operate in every legal system. Clearly, the Convention does not prohibit such presumptions in principle. It does, however, require the Contracting States to remain within certain limits in this respect as regards criminal law.”
[37] In Zuma above n 28 at para 41, Kentridge AJ emphasised that the effect of his judgment was not to invalidate every legal presumption reversing the onus of proof, since some presumptions,
“may be justifiable as being rational in themselves, requiring an accused person to prove only facts to which he or she has easy access, and which it would be unreasonable to expect the prosecution to disprove . . . Or there may be presumptions which are necessary if certain offences are to be effectively prosecuted, and the State is able to show that for good reason it cannot be expected to produce the evidence itself.”

This statement was cited in Mbatha above n 7 at para 15 and Coetzee above n 34 at para 13. In Coetzee, Langa J noted at para 13 that,
“[t]here is no doubt a pressing social need for the effective prosecution of crime. Kentridge AJ, speaking for the Court in Zuma, noted that reasonable presumptions may be required by the prosecution, in relation to certain categories of offences, to assist in this task.” (Footnote omitted)
[38] The approach developed in the concurring judgment of Harlan J in In re Winship [1970] USSC 77; 397 US 358 (1970) is instructive. He noted that the standard of proof in judicial proceedings represented an attempt to instruct the fact-finder concerning the degree of confidence society thought should be had in the correctness of the factual conclusions for a particular type of adjudication. In criminal cases, the social disutility of convicting an innocent person was not viewed as equivalent to the disutility of convicting someone who was innocent. He adopted the following observation by Brennan J in Speiser v Randall [1958] USSC 154; 357 US 513 (1958) at 525-6:
“There is always in litigation a margin of error, representing error in factfinding, which both parties must take into account. Where one party has at stake an interest of transcending value - as a criminal defendant his liberty - this margin of error is reduced as to him by the process of placing on the other party the burden of producing a sufficiency of proof in the first instance, and of persuading the factfinder at the conclusion of the trial of his guilt beyond a reasonable doubt.”

Although Winship dealt with the question whether guilt at juvenile delinquency proceedings should be proved beyond a reasonable doubt or merely on a balance of probabilities, the balancing approach adopted by Harlan J is helpful in applying limitations analysis to a reverse onus. The issue is essentially the same: the degree of risk of erroneous conviction considered to be acceptable in an open and democratic society, bearing in mind all the circumstances.
[39] See the foreign cases referred to by Kentridge AJ in Coetzee above n 34 at para 96.
[40] Coetzee above n 34 at para 217-8.
[41] For example, the provisions of section 237 of the Criminal Procedure Act with regard to evidence on a charge of bigamy.
[42] In Meaker above n 36 at 1057J, Cameron J concluded that section 130(1) of the Road Traffic Act, 29 of 1989, which provided:
“Where in any prosecution under the common law relating to the driving of a vehicle on a public road, or under this Act, it is material to prove who was the driver of a vehicle, it shall be presumed, until the contrary is proved, that such vehicle was driven by the owner thereof”,
was not unconstitutional, since the presumption created by the section was
“ . . . an eminently reasonable device, which accords with practical common sense and in its application produces equitable results.”
[43] S v Mahlinza 1967 (1) SA 408 (A) at 419; S v Trickett 1973 (3) SA 526 (T) at 532. For a critical view see Paizes above n 35 SA Journal of Criminal Justice at 412 at footnote 3. In Canadian law, a provision of the Criminal Code which enacted a presumption of sanity and required insanity to be proved by the accused was held to infringe section 11(d) of the Charter, but was upheld under the limitation inquiry in section 1; R v Chaulk (1990) 1 CRR (2d) 1, followed in R v Ratti (1991) 1 SCR 68 and R v Romeo (1991) 1 SCR 86. In R v Chaulk, Lamer CJC noted that an accommodation of three important societal interests was involved: avoiding a virtually impossible burden on the Crown; convicting the guilty; and acquitting those who truly lacked the capacity for criminal intent.
[44] S v Makwanyane and another [1995] ZACC 3; 1995 (3) SA 391 (CC); 1995 (6) BCLR 665 (CC) at para 104.
[45] Id.
[46] See Dlamini above n 6 at para 68.
[47] See Mistry v Interim Medical and Dental Council of South Africa and Others [1998] ZACC 10; 1998 (4) SA 1127 (CC); 1998 (7) BCLR 880 (CC) at para 30.
[48] Osman above n 15 at para 17.
[49] Id.
[50] See S v Bequinot [1996] ZACC 21; 1997 (2) SA 887 (CC), 1996 (12) BCLR 1588 (CC) at para 12:
“The receipt of stolen goods is a vital link in the chain of gainful disposal of the spoils of criminality. It is, of course, also a powerful incentive to such criminality and statutory devices aimed at facilitating the successful apprehension and prosecution of receivers of stolen property, such as section 37 clearly is, cannot lightly be invalidated”.
[51] Above n 6 at para 68.
[52] Coetzee above n 34 at para 220.
[53] Mbatha above n 7 at para 19.
[54] Mello above n 30 at para 9; Ntsele above n 30 at para 4; Mbatha above n 7 at para 19; Bhulwana above n 30 at para 18.
[55] See discussion above at para 17.
[56] See National Coalition for Gay and Lesbian Equality and others v Minister of Home Affairs and others [1999] ZACC 17; 2000 (1) BCLR 39 (CC) at para 75.
[57] Above n 30.
[58] Above n 56.
[59] Section 35 (2) states,
“No law which limits any of the rights entrenched in this Chapter, shall be constitutionally invalid solely by reason of the fact that the wording used prima facie exceeds the limits imposed in this Chapter, provided such a law is reasonably capable of a more restricted interpretation which does not exceed such limits, in which event such law shall be construed as having a meaning in accordance with the said more restricted interpretation.”
[60] Section 232(3) states,
“No law shall be constitutionally invalid solely by reason of the fact that the wording used is prima facie capable of an interpretation which is inconsistent with a provision of this Constitution, provided such a law is reasonably capable of a more restricted interpretation which is not inconsistent with any such provision, in which event such law shall be construed as having a meaning in accordance with the said more restricted interpretation.”
[61] Section 98 states,
“(5) In the event of the Constitutional Court finding that any law or any provision thereof is inconsistent with this Constitution, it shall declare such law or provision invalid to the extent of its inconsistency: Provided that the Constitutional Court may, in the interests of justice and good government, require Parliament or any other competent authority, within a period specified by the Court, to correct the defect in the law or provision, which shall then remain in force pending correction or the expiry of the period so specified.”

(6) Unless the Constitutional Court in the interests of justice and good government orders otherwise, and save to the extent that it so orders, the declaration of invalidity of a law or a provision thereof—

(a) existing at the commencement of this Constitution, shall not invalidate anything done or permitted in terms thereof before the coming into effect of such declaration of invalidity; or

(b) passed after such commencement, shall invalidate everything done or permitted in terms thereof.

(7) In the event of the Constitutional Court declaring an executive or administrative act or conduct or threatened executive or administrative act or conduct of an organ of state to be unconstitutional, it may order the relevant organ of state to refrain from such act or conduct, or, subject to such conditions and within such time as may be specified by it, to correct such act or conduct in accordance with this Constitution.
[62] Above n 30 at paras 26 - 28.
[63] Above n 30.
[64] Above n 56.
[65] Id at para 65.
[66] Id at para 70.
[1] Section 3 of the Stock Theft Act, 57 of 1959, contains a similar provision to section 37 but relates only to possession of stolen stock or produce. Section 37, read with that section, makes it clear that possession of stock or produce is governed by the provisions of the Stock Theft Act and not by section 37.
[2] See S v Moller [1990] ZASCA 62; 1990 (3) SA 876 (A) at 888 A - B; S v Ghoor 1969 (2) SA 555 (A) at 557.
[3] See S v Zuma and Others [1995] ZACC 1; 1995 (2) SA 642 (CC); 1995 (4) BCLR 401 (CC) at para 33; S v Bhulwana; S v Gwadiso [1995] ZACC 11; 1996 (1) SA 388 (CC); 1995 (12) BCLR 1579 (CC) at para 15; S v Mbatha; S v Prinsloo [1996] ZACC 1; 1996 (2) SA 464 (CC); 1996 (3) BCLR 293 (CC) at paras 9 - 12; S v Julies [1996] ZACC 14; 1996 (4) SA 313 (CC); 1996 (7) BCLR 899 (CC) at para 3; Scagell and Others v Attorney-General, Western Cape, and Others [1996] ZACC 18; 1997 (2) SA 368 (CC); 1996 (11) BCLR 1446 (CC) at para 7; S v Coetzee and Others [1997] ZACC 2; 1997 (3) SA 527 (CC); 1997 (4) BCLR 437 (CC) at para 8; S v Ntsele [1997] ZACC 14; 1997 (2) SACR 740 (CC); 1997 (11) BCLR 1543 (CC) at para 3; S v Mello and Another [1998] ZACC 7; 1998 (3) SA 712 (CC); 1998 (7) BCLR 908 (CC) at paras 4 - 6; S v Manyonyo 1999 (12) BCLR 1438 (CC) at para 11.
[4] [1995] ZACC 3; 1995 (3) SA 391 (CC); 1995 (6) BCLR 665 (CC) at para 104.
[5] 1999 (1) SA 6 (CC); 1998 (12) BCLR 1517 (CC) at paras 33 - 5.
[6] S v Zuma above n 3 at para 33; S v Bhulwana; S v Gwadiso above n 3 at paras 10, 15 - 6.
[7] See for example, In re Winship [1970] USSC 77; 397 US 358 (1970) at 363 - 4.
[8] In this regard, it is of significance that the Canadian Supreme Court has on several occasions held reverse onuses to be justified limitations of the presumption of innocence. See for example, R v Whyte [1989] 51 DLR (4th) 481; R v Chaulk [1991] 1 CRR (2d) 1; R v Wholesale Travel Group Inc [1992] 84 DLR (4th) 161; and R v Downey [1992] 90 DLR (4th) 449.
[9] See for example, S v Bhulwana; S v Gwadiso above n 3 at para 18.
1[0] S v Ghoor above n 2 at 557, citing with approval Trollip J in S v Kaplin and Others 1964 (4) SA 355 (T) at 358.
[1]1 See for example, Demmers v Wyllie and Others 1978 (4) SA 619 (D) at 629 per Didcott J; S v Robson; S v Hattingh 1991 (3) SA 322 (W) per Kriegler J at 333; S v Ngema 1992 (2) SACR 651 (D) per Hugo J at 655 - 7.
1[2] See for example, De Wet and Swanepoel Strafreg 4 ed (Butterworth, Durban 1985) at 156 - 63. Others have noted the difficulties but have not suggested an abandonment of the objective test. See Snyman Criminal Law 2 ed (Butterworths, Durban 1989) at 230 - 5; Milton South African Criminal Law and Procedure Vol II 3 ed (Juta, Cape Town 1996) at 385 - 90.
1[3] For a philosophical justification of the legal requirement of objectively reasonable conduct, tempered by the particular circumstances in which the accused found himself or herself, see Honoré Responsibility and Fault (Hart, Oxford - Portland 1999) at 34 - 7 and 122.
1[4] 1933 AD 269 at 272.
1[5] See for example, S v Mini 1963 (3) SA 188 (A) at 196 G - H; S v Bernardus 1965 (3) SA 287 (A) at 300 E - F; S v Burger 1975 (4) SA 877 (A) at 879 D; S v Bochris Investments (Pty) Ltd and Another 1988 (1) SA 861 (A) at 865 G - I.
1[6] 1976 (2) SA 921 (A) at 928 C - E.
1[7] Above n 2 at 559 C. (our translation)
1[8] Above n 3.
1[9] Above n 3.
2[0] Compare the discussion in Coetzee above n 3 at paras 86 and 93 per Kentridge AJ and paras 159 - 60 per O’Regan J.
2[1] S v Parrow 1973 (1) SA 603 (A) at 604 B - F; S v Skweyiya [1984] ZASCA 96; 1984 (4) SA 712 (A) at 715 - 6.
[2]2 As the court inferred in S v Skweyiya above n 21.
2[3] [1996] ZACC 21; 1997 (2) SA 887 (CC); 1996 (12) BCLR 1588 (CC) at para 12.
2[4] See S v Meaker 1998 (2) SACR 73 (W); 1998 (8) BCLR 1038 (W). This was manifestly not the case in many of the earlier cases concerning reverse onuses considered by this Court. See for example, S v Bhulwana; S v Gwadiso above n 3 at para 23; S v Mbatha; S v Prinsloo above n 3 at para 22.
2[5] Above n 2 at 886 F - G.
2[6] Above n 3 at para 41.
2[7] Id.
2[8] Illinois State Board of Elections v Socialist Workers Party et al [1979] USSC 27; 440 US 173 (1979) at 188 - 9.
2[9] See S v Coetzee above n 3 at para 195.
3[0] Id. See also R v Wholesale Travel Group Inc above n 8 at 234 - 5 per Iacobucci J.
3[1] Honoré above n 13 at 125.