LLB (Hons) (Malawi) LLM PhD (Cape Town)
Postdoctoral Research Fellow, Department of Public Law, University of Cape Town, South Africa
Edition: AHRLJ Volume 17 No 2 2017
Pages: 436 - 459
Citation: E Gumboh ‘Correct but wrong: A critical analysis of recent jurisprudence on the proper test for release pending appeal applications in Malawi’(2017) 17 African Human Rights Law Journal 436-459
Download article in PDF
In Kumuwa v Republic, the High Court of Malawi held that the Constitution did not distinguish between release before and after conviction. Relying on section 42 of the Constitution, which provides for the right to be released from detention with or without bail unless the interests of justice require otherwise, the Court reasoned that there was one test for release: the interests of justice. This position was rejected by the same Court in Uche v Republic. The Malawi Supreme Court of Appeal in Kumwembe v Republic upheld the correctness of Uche, and reiterated that the Constitution had not changed the traditional test of exceptional circumstances for release pending appeal. This article argues that Kumuwa rightly identified the test as the interests of justice, but erred in finding that the constitutional premise was the right in section 42. Consequently, Uche and Kumwembe miss the point when they discard the test advanced in Kumuwa solely on this faulty constitutional premise. The article aims to discuss the Court’s reasoning and evaluate it in light of the law and jurisprudence on the proper test for granting release pending appeal.
Malawian courts have always distinguished between release pending trial and release pending appeal. The test for release pending trial is derived from the right of all accused persons to be released from detention with or without bail unless the interests of justice require otherwise.1 This means that before conviction, release is automatic unless the state discharges its onus of showing that the interests of justice require the further detention of the suspect. Accordingly, an accused need not apply for release, but a court must, on its own motion, release the accused. However, after conviction, the legal status of an offender changes and the consideration for release has been held to hinge on whether there are exceptional and unusual circumstances warranting release pending appeal, with the onus at this stage cast on the offender to persuade a court to release him or her from custody.
In Kumuwa v Republic,2 the High Court of Malawi held that this approach to release an applicant pending appeal was fundamentally flawed as it did not take cognisance of the Constitution, particularly section 42, which guarantees the right to be released from detention unless the interests of justice require otherwise. The Court reasoned that the Constitution, therefore, dictated the interests of justice as the proper test for release both before and after conviction. This decision is significant, as it not only upsets the settled approach to release pending appeal, but also raises complex issues about the onus of proof in applications for release pending appeal and the legal status of a trial court’s sentence. Kumuwa was later rejected by the High Court in Uche v Republic3 and by the Malawi Supreme Court of Appeal (MSCA) in Kumwembe v Republic.4 Both decisions reiterated that the Constitution had not changed the traditional test of exceptional circumstances for release pending appeal.
The article seeks to highlight the soundness of the Court’s reasoning in Kumuwa, and to assess whether the conclusion reached is legally tenable. It also examines the premise on which Uche and Kumwembe differed from Kumuwa. The article argues that Kumuwa rightly identified the proper test for granting release pending appeal as the interests of justice. However, it erred in its conclusion that the justification for this test springs from section 42. Consequently, Uche and Kumwembe miss the point when they discard the test advanced in Kumuwa solely on this faulty constitutional premise. It is apposite to begin with a brief overview of case law in Malawi on the release of an applicant pending appeal leading up to Kumuwa.
There is no statutory guidance on the test applicable to release pending appeal. Section 118 of the Criminal Procedure and Evidence Code (CPEC)5 and the Bail Guidelines Act6 are limited to pre-trial release. However, there is no gap in the law because judicial precedent provides ample instruction on how a court must approach the question of release pending appeal. In the 1971 case of Pandirker v Republic,7 Justice Chatsika held that release pending appeal should be granted only when the accused shows that there are exceptional circumstances warranting it.8 Over 20 years later, the judge reinforced this position in Chihana v Republic:9
In an application for bail pending an appeal, it has to be borne in mind that, upon conviction, the applicant lost his freedom of movement. In essence, conviction is followed by punishment. The authorities have a duty to restrict, as one of the forms of the punishment, his freedom, on the basis of the conviction. He is no longer a free man. Therefore, in order to grant freedom to such a person whose fundamental freedom has been lost by the conviction, there must exist some ‘exceptional and unusual circumstances’. In other words, the case must be so exceptional and unusual that having regard to all the circumstances surrounding it, the court will be justified in overlooking the order for his imprisonment and make a counter order that he be released, at least until his appeal has been determined.
This test has survived the constitutional era as confirmed in several cases by the Malawi Supreme Court of Appeal (MSCA). In Suleman v Republic,10 Justice of Appeal Tembo stated that the test espoused in Chihana was an ‘authoritative enunciation’ of the law applicable to applications for release pending appeal. The threshold for ‘exceptional and unusual circumstances’ is high. In Kaliati v Republic, it was held that release pending appeal
is the exception rather than the norm. Put slightly differently, the law presumes the sentence passed on a convict by any court of law to be right, and to remain right and deserving to be undergone until such time as a superior court has had a chance to look at it and to pronounce it otherwise. The law accordingly only allows interference with such sentence by way of staying its operation, whilst awaiting an appeal if and only if the convict succeeds to dispel the applicable presumption by bringing to the attention of the court he is applying to for bail, circumstances that are visibly and convincingly ‘special’ or ‘exceptional’.11
What amounts to an ‘exceptional circumstance’ has not been the subject of much legal debate in Malawi. The difficulty of precisely defining this phrase has correctly been described as a ‘legal headache’.12 It stands to reason that if the test is ‘exceptional and unusual circumstances’, then it is impossible to have an exhaustive list of exceptional circumstances as each case will turn on its own facts, and a court will have to exercise its discretion in deciding whether the test is met or not. Indeed,
[i]f the court takes into account factors already on the case roll they might be exhaustive, but with the changing world, they may not. There is the likelihood of new exceptional circumstances resurfacing as and when social and mankind behavioural changes occur in the present world.13
Case law indicates that exceptional circumstances include the following: the possibility that the appeal will be successful;14 a missing court record;15 delays in the appeal process, for instance, due to the preparation of the court record;16 poor health of the accused;17 a real possibility that the accused will raise money to pay a fine while on bail;18 the unavailability or inaccuracy of the court record;19 the likelihood that the appeal cannot be concluded within a reasonably short time;20 the risk that the sentence will be served by the time the appeal is heard;21 and whether a conviction on a lesser offence would be competent.22 Unless ‘exceptional’, the fact that the accused person’s family will suffer hardship as a result of imprisonment does not constitute an exceptional circumstance.23 Good character of the applicant is also irrelevant.24
It is worth noting that the courts have not been consistent in their approach to these factors. While it is understandable that a factor may be considered exceptional in one case and not in another in view of the facts of the case at hand, it is disturbing when a factor is accepted as sufficient in one case and rejected out of hand in another. This is well exemplified by factors such as ill-health,25 a missing court record26 and the likelihood of a sentence being served before the appeal is concluded.27
Exceptional circumstances are not examined in isolation. Once an exceptional circumstance has been proved, some considerations applicable to release pending trial come into play mutatis mutandis. The primary consideration is whether the accused will attend the appeal hearings.28 Release, therefore, will not be granted unless the court is satisfied that the accused will not abscond. In deciding this question, a court may take into account the fact that the accused was earlier released on and honoured his bail pending trial.
However, as rightly observed by the MSCA, the jurisprudence indicates ‘that generally courts are loath, and indeed very reluctant, to admit new instances or categories of exceptional and unusual circumstances outside and beyond the parameters’ of the likelihood that the appeal will succeed or because of the risk that the sentence will have been served by the time the appeal is heard.29 In fact, courts tend to emphasise the existence of exceptional and unusual circumstances as the requirement and not the interests of justice. In Chihana, Justice Chatsika went a step further to add a concurrency requirement:30
It seems that where it appears, prima facie, that the appeal is likely to be successful or where there is a risk that the sentence will be served by the time the appeal will be heard, the test will have been satisfied. I think that the two factors must exist concurrently in order for the condition to be satisfied.
This was reiterated by Justice Unyolo in Kamaliza v Republic.31
The major problem with this reasoning is that it fails to recognise that the likelihood of success and undue delay are only examples of exceptional and unusual circumstances.32 This raises many difficulties in determining how to deal with various factors in release pending appeal applications. Indeed, the case law suggests that these two factors often are overemphasised, such that other potential circumstances are overlooked. In Nathebe it was held that ‘[t]he only matter to consider is whether prima facie there is a likelihood that the appeal here will succeed’.33 This leads to release being denied where it should have been granted, such as where the court record is missing.
A second problem with the concurrency requirement is that it inhibits a court’s ability to independently deal with the likelihood of the sentence being served, yet this factor goes to the root of release pending appeal which is to avoid the execution of a sentence of imprisonment until it has been confirmed by a higher court. Moreover, the success of the appeal is obviously not considered fully during the bail application.34 Here again, requiring that the likelihood of success should be the overriding consideration becomes problematic.
In view of these challenges arising from the concurrency requirement, the decision of the MSCA in Suleman should be welcomed to the extent that it clarifies the fact that this requirement is erroneous. However, the decision is a source of greater dissatisfaction about what the meaning of the exceptional and unusual circumstance test is. The Court in Suleman correctly pointed out that the concurrency requirement was not supported by case law, let alone the case law cited in Chihana itself:35
It is submitted that the expression ‘exceptional and unusual circumstances’ in that context [of applications for release pending appeal] means circumstances where, on the one hand, it appears prima facie that the appeal is likely to be successful or, on the other hand, where there is a risk that the sentence will have been served by the time the appeal is heard. Where either of these factors exist, the court may grant an application for bail pending appeal. Thus, it must be fully and clearly appreciated that the existence of either of those factors, standing on its own, sufficiently constitutes a requisite condition precedent for the granting of such application by the courts.
The Court then, quite unexpectedly, went on to hold that there were but two exceptional and unusual circumstances:36
To begin with, it must be pointed out that [an application for release pending appeal] ought to be considered and determined on the basis of the existence of exceptional and unusual circumstances in the case; that is to say in the light, and within the parameters, of either of the factors, which courts have prescribed as conditions precedent to the granting of application for bail pending appeal. In the view of case authorities on the point, all the arguments made in support of the application, on the one hand, on the basis of the applicant’s status and financial commitments, as a very successful businessman and, on the other hand, respecting his ill-health are matters which do not constitute exceptional and unusual circumstances. It is so held in that regard. Whilst heeding the above mentioned caution in the matter, I must nonetheless ask myself the following questions. Thus, having read the judgment in the case, including the relevant court record, do I hold the impression that the appellant’s appeal has a fair chance of being successful? If my response is not in the affirmative, and only then alternatively, do I hold the view that the sentence will have expired by the time the appeal shall have been determined?
Granted, this disjunctive approach in Suleman is a slightly more flexible approach than the concurrency requirement in Chihana and Kamaliza, and resolves some of the challenges raised earlier. However, the approach in Suleman has its setbacks. For example, it is still restrictive in that it limits the scope of exceptional and unusual circumstances to the likelihood of success of an appeal and the risk that the sentence may have been served by the time the appeal is heard. Suleman is also not wholly in tandem with precedent. The Court was clear that courts are reluctant to extend the scope of exceptional and unusual circumstances beyond these two factors.37 While it is true that most of the factors considered as exceptional or unusual by the courts relate to the likelihood of a successful appeal and the risk of the sentence being served before the appeal is determined, there are other factors that have been held to satisfy the test.38 Hence, the conclusion in Suleman is regrettable since there are many instances where the interests of justice may require that release pending appeal should be granted. A further fault in Suleman is that it is irreconcilable with the fact noted earlier, that it is impossible to have an exhaustive list of exceptional and unusual circumstances. Actually, it flies in the face of the very meaning of the notion of ‘unusual and exceptional circumstances’ in everyday language to hold that the circumstances in fact are only two. Such a conclusion effectively means that the circumstances are no longer ‘exceptional’ or ‘unusual’ in any sense of the word. According to the Oxford English dictionary, ‘exceptional’ means ‘very unusual’, while ‘unusual’ entails ‘different from what is usual or normal’. It follows from this that once the circumstances are reduced to two, they lose the character of being exceptional or unusual.
Given this background, it was not unexpected that the High Court sought to set the record straight on the proper test for release pending appeal in Kumuwa. Whether the Court got it right this time around is considered below.
The three accused were serving a 10-year default sentence for the theft of cattle. In their application for release pending appeal, the appellants argued that the appeal was likely to succeed and that there was a possibility that the appeal would be delayed in view of the backlog of cases to be heard by the Court. While the Court raised and dealt with a number of issues in its judgment, it is the finding on the application for release that is of importance for present purposes.
Justice Mwaungulu criticised the test of exceptional and unusual circumstances as inadequate and unconstitutional. He held that ‘there are bound to be cases of usual and unexceptional nature where, pending an appeal, the court may grant bail or stay execution of a sentence’.39 The problem with the jurisprudence on release pending appeal, according to the judge, is that it transposes what should ordinarily be considered factors to the status of a test.40 The Court held that the real test for all release applications before or after conviction must be the interests of justice.41 It stressed that the likelihood of success on appeal and the risk that the sentence may be served before an appeal is heard should only be factors to be taken into account in determining whether the interests of justice require that release should be granted.42 The Court was at pains to demonstrate ‘that the prospect of success of a case is just a factor that informs the court where interests of justice lie’.43 It ruled that ‘the much vaunted “exceptional and unusual circumstances” would be totally inadequate or unnecessary’ in certain cases where the circumstances were ‘unexceptional and usual’, yet the interests of justice tilted in favour of granting release pending appeal.44 It pointed out that the problem would be that any test formulated so as to encompass all possible factors would be ‘very pregnant as to be no test at all’. Moreover, the Court added, the test of unusual or exceptional circumstances fetters the discretion of a court as to the factors warranting release on release pending appeal. It observed that the discretion to grant release pending appeal would be unduly fettered if a factor such as the existence of exceptional or unusual circumstances is elevated to a test instead of complementing the discretion in the balancing of justice. The Court concluded that the dominant principle in release pending appeal applications must be the interests of justice as informed by various factors, including the prospect of success of the appeal. It explained that the interests of justice would vary depending on the stage of the criminal process. Thus, at the post-sentence stage, the presumption of innocence and interference with witnesses ordinarily play no role; only the sentence, prescribed and imposed, the offender, the victim and the public interest would be relevant.
Justice Mwaungulu noted that the courts had continued to apply the same ‘unusual and exceptional circumstances’ test used before the adoption of the democratic Constitution in 1994 without considering the implications on bail of this Constitution. He argued that the Constitution mandated ‘one uniform and unified test for granting bail in all cases, before or after conviction, namely, the interest[s] of justice’. The Court’s reasoning mainly turned on the wording of section 42, particularly section 42(2)(e). It is necessary to quote section 42 extensively as it is key to contextualising the decision of the Court:
(b) to be held under conditions consistent with human dignity, which shall include at least the provision of reading and writing materials, adequate nutrition and medical treatment at the expense of the state;
(c) to consult confidentially with a legal practitioner of his or her choice, to be informed of this right promptly and, where the interests of justice so require, to be provided with the services of a legal practitioner by the state;
(d) to be given the means and opportunity to communicate with, and to be visited by, his or her spouse, partner, next-of-kin, relative, religious counsellor and a medical practitioner of his or her choice;
(b) as soon as it is reasonably possible, but not later than 48 hours after the arrest, or if the period of 48 hours expires outside ordinary court hours or on a day which is not a court day, the first court day after such expiry, to be brought before an independent and impartial court of law and to be charged or to be informed of the reason for his or her further detention, failing which he or she shall be released;
(v) to be represented by a legal practitioner of his or her choice or, where it is required in the interests of justice, to be provided with legal representation at the expense of the State, and to be informed of these rights;
(vi) not to be convicted of an offence in respect of any act or omission which was not an offence at the time when the act was committed or omitted to be done, and not to be sentenced to a more severe punishment than that which was applicable when the offence was committed;
(vii) not to be prosecuted again for a criminal act or omission of which he or she has previously been convicted or acquitted, save upon the order of a superior court in the course of an appeal or review proceedings relating to that conviction or acquittal;
ix) to be tried in a language which he or she understands or, failing this, to have the proceedings interpreted to him or her, at the expense of the state, into a language which he or she understands; and
According to the Court, section 42(2)(e) provides ‘a general right not restricted to bail before or after conviction’.45 It stated that a careful reading of section 42(2) demonstrates that the interests of justice are the dominant test for all bail applications, before and after conviction.46 The Court explained that the rights in section 42 were cumulative and that some rights relating to detention in section 42(1) applies to all detainees regardless of their status, that is, including suspects and sentenced offenders alike.47 To be sure, ‘a sentenced offender has the right to bail which a fortiori is bail pending appeal’.48 Cementing this interpretation, Justice Mwaungulu then went on to clarify that the reference to an ‘accused’ in the opening statement of section 42(2) was not limited to suspects. Rather, it
includes sentenced offenders, those convicted and sentenced of a crime. Section 42(2) in its introduction gives the rights to the accused ... in relation to the right to a fair trial in section 42(2)(f), ‘as an accused person’ the ‘accused’ has rights like ‘to have recourse by way of appeal or review to a higher court than the court of first instance’ (section 42(2)(f)(viii)) and to be ‘sentenced within a reasonable time after conviction’ (section 42(2)(f)(viii)) which are consistent with the suggestion that as convicted or sentenced offenders they are accused persons who must enjoy the right to bail under section 42(2)(e) of the Constitution. Moreover, section 42(2)(e) by its wording does not limit the right to bail by qualifiers like ‘before’ or ‘after’ conviction; the right is unfettered and unqualified in any respect.49
The Court further stated that sentenced offenders must also enjoy the right to challenge the lawfulness of their detention and to be released if it is found to be unlawful as provided in sections 42(1)(e) and (f).50 However, it emphasised that, although the interests of justice is the dominant principle that should govern applications for release pending appeal, there could be other principles.51 What is critical, the Court continued, is that a court considers all relevant factors and circumstances ‘arising latently or patently in the application’.52 It warned that a court should not
orchestrate a minor factor or undermine an important consideration. These factors or circumstance[s], therefore, must not limit the discretion; they must augment and complement the discretion in balancing justice.53
To further buttress its conclusions, the Court invoked section 11(2) of the Constitution which provides that in ‘interpreting the provisions of this Constitution a court of law shall where applicable, have regard to current ... comparable foreign case law’. Justice Mwaungulu asserted that in as far as ‘bail provisions’ are concerned, ‘South African decisions are like [Malawian decisions]. There the test for bail before conviction and pending appeal has been unified.’ He then cited Smith v S:54
All bail applications are governed by the same principle, but there are differences in emphasis in cases of bail pending an appeal from cases of bail prior to conviction ... [T]he overriding consideration remains always the right of an accused person in terms of section 25(2)(d) of the Constitution to be released on bail unless the interests of justice require otherwise.
Turning to Zimbabwe, the Court discounted the High Court’s decision in Manyange v The State,55 which held that release pending appeal rested on whether there were reasonable prospects of success on appeal and that the release would not endanger the interests of justice.56 In disagreement, Justice Mwaungulu described Manyange as being ‘caught between modernity and tradition, combining a circumstance or factor with the test’.57 He also faulted Manyange for citing the pre-Constitution South African case of R v Mthembu.58
Fortified in its interpretation of section 42(2)(e), Kumuwa concluded that the constitutional right to be released extended to sentenced offenders and mandated the interests of justice as the proper test for release pending appeal applications. On this understanding, the Court questioned why the Bail Guidelines Act was limited to release before conviction in terms of section 118 of the CPEC. It concluded that the factors listed in this Act should also inform the approach to post-conviction release except in cases where the death penalty had been imposed.
Inspired by this new approach in Kumuwa, the appellants in Uche argued that an accused no longer needed to prove exceptional circumstances to be released pending appeal. Rejecting this submission, the Court was of the view that the reasoning in Kumuwa was ‘quite interesting’59 but ‘erroneous’.60 Justice Kachale observed that there was inadequate justification for the conclusion that ‘sentenced prisoners’ were ‘accused persons’ within the meaning of section 42(2).61 In his view, this was incompatible with the legal status of a convicted and sentenced person since, at that stage, any allegations against him had graduated into factual findings rendering him outside the ambit of persons envisaged in section 42(2).62 The Court thus upheld ‘the historical position that post-conviction and sentence, the burden on the convict seeking bail is necessarily quite onerous’.63
The MSCA in Kumwembe had the opportunity to ‘clarify the law’ governing release pending appeal by considering the correctness of Kumuwa. The appellant in this case had applied for release pending appeal against his conviction and four months’ sentence for contempt of court. Reasoning that the test for stays and release pending appeal was the same, the case centred on Kumuwa and its impact on the applicable test. As in the case of the applicant in Uche, Kumwembe argued that it was no longer necessary to show exceptional or unusual circumstances before granting a stay, and that all that was needed was for him to show that it was in the interests of justice that he be granted the stays prayed for.
Justice of Appeal Chikopa observed that the law on the test for granting release pending appeal had been clear until Kumuwa cast doubt on it. According to the Court, Kumuwa ‘seemed to be saying that the standard for granting a stay or bail pending appeal was the same as that for bail pending trial’.64 Like the High Court in Uche, the MSCA restricted itself to the question of whether section 42(2)(e) extended to convicted persons, and concluded that it was ‘obvious’65 that this was not the case. Capitalising on the latter part of the opening paragraph of section 42(2), which refers to a person ‘arrested for, or accused of the alleged commission of an offence’, the Court stressed that the provision was only applicable to suspects and excluded convicts.66 It explained that a detained suspect was also entitled to rights under section 42(1) by virtue of being a detainee.67 Treading carefully, the Court continued:68
There are obvious overlaps in the identity of the persons envisaged in section 42(1) and (2) and the manner in which the Constitution thought we should deal with them. There is no denying however that they are clearly two distinct persons. In subsection (1) is a detainee who might also be a convict. In subsection (2) is a suspect who might also be a detainee but is certainly not a convict. There is also no denying that our Constitution in some important respects treats these two persons differently. For instance, whereas the suspect has all the rights of a detainee, the reverse is not true. Another instance is section 42(2)(e). It grants the detained suspect the right to be released from detention, with or without bail unless the interests of justice require otherwise. There is no identical provision for the detainees in [section 42(1)].
‘[I]t is therefore to err’, the Court concluded, ‘to think that the Constitution conferred the right in section 42(2)(e) on sentenced prisoners [convicts] or to seek, via judicial opinion, to arrogate such right to them’. The Court, therefore, concurred with Uche. It endorsed the traditional approach to release pending appeal as explicated in Pandirker and stressed that the factors listed therein were examples.69 Importantly, the MSCA added that ‘[t]he list of examples is most likely longer’.70 The Court then held that a convict bore the burden ‘to show exceptional, special and unusual circumstances’ before release is granted.71 Accordingly, it found that in the instant case the question was whether the applicant had placed ‘sufficiently special, unusual and exceptional circumstances’ before the Court to warrant the stays sought.72
It is important to mention that Kumwembe is not correct by asserting that the test for release pending appeal had been clear until Kumuwa. As explained earlier, the jurisprudence did not speak with one voice on the meaning of exceptional circumstances. However, both Uche and Kumwembe are correct in finding that sentenced prisoners are not accused persons within the meaning of section 42(2). However, neither case addresses the interests of justice test argument, that is, the contention in Kumuwa that the courts have elevated factors relevant to release pending appeal to tests and that ‘most circumstances or factors inform the interests of justice, they are not tests’ separately.
The impetus to devise a new test for release pending appeal was that the present approach is too restrictive. This is understandable, given that courts have traditionally limited exceptional circumstances. In fact, the MSCA has previously held that, in effect, there are only two exceptional circumstances, namely, the prospects of success of the appeal and the risk that the sentence will have been served by the time the appeal is heard.73 Consequently, courts have denied release in cases that raise other circumstances, such as missing court records, making it impossible to prosecute an appeal. Even ill health has been discounted as an exceptional circumstance. While some cases have been more generous and have granted release on factors outside the now seemingly closed list, the majority of cases have stuck to the traditional approach. Kumuwa, therefore, is right in criticising the approach to release pending appeal and recasting the inquiry as one of justice.
It stands to reason that the proper test, indeed, is the interests of justice as determined by the circumstances of the case. This is what the courts have stated in several cases though ultimately relegated to the traditional restrictive approach. For instance, in Republic v Ndomondo, the MSCA noted that the grant of an order of stay was ultimately ‘a matter for judicial discretion based on the circumstances of the case and the justice of the matter’.74 As observed in Kumwembe, the factors listed in Pandirker are but ‘examples of exceptional and unusual ... The list of examples is most likely longer.’ One might go further to state that the list is obviously longer. Surely, if the listed factors are examples, the list is not exhaustive as there must be other factors that are not listed? Interestingly, a perusal of some cases indicates that courts have stressed that case law only provides examples; the fact that some cases recognise factors beyond those in Pandirker lends credence to this point.75 The law then is that the test remains the interests of justice as informed by a wide range of factors. An attempt to close the list of circumstances that would justify the granting of release pending appeal lies at the heart of the problem that Kumuwa sought to rectify. As noted above, at some point it was even thought that the Pandirker factors must exist simultaneously, a proposition which was fortunately corrected but, unfortunately, continues to be endorsed in some cases.
Therefore, it has not always been obvious in the courts’ approach to release a convict pending appeal that the list of factors that may justify release pending appeal remains subject only to the interests of justice and logic. As noted in Kumuwa, the courts have confused the test and the factors that must inform it. Instead of looking at exceptional circumstances as factors, the focus has been placed on the circumstances to such an extent that the highest court of the land has reduced these factors to two. The very nature of the interests of justice is such that the list of factors that can inform a court where the interests lie cannot be closed. Indeed, if only two factors can qualify as exceptional circumstances, they are no longer exceptional. This approach leads to a situation where other factors are ignored, resulting in injustice. For example, a missing court record would disadvantage an offender since the prospects of the appeal succeeding cannot be evaluated. Moreover, other factors, such as ill health, would and are excluded. Justice Mwaungulu, therefore, is right when he holds that in pursuance of the interests of justice, a court should be open to take account of a myriad of factors in determining an application for release pending appeal. Speaking of the factors laid out in the Bail Guidelines Act which is limited to release pending trial, Kumuwa aptly notes that the Act
is dominated by one fundamental consideration: the interest(s) of justice. On this foundation is built factors that may be [taken] into account either generally or in the pursuance of the interest(s) of justice. The factors or circumstances taken into account are not a test or requirement for granting bail. The factors are neutral and can, therefore, influence the result one way or the other. They are integral to the question whether, in the interests of justice, bail should be granted.76
Such an approach to release pending appeal would go a long way in ensuring that the interests of justice remain paramount, and that courts do not relegate the inquiry to one of whether the appeal is likely to succeed or the sentence may have been served before the appeal is determined. This is the common law position that needs no further constitutional backing.
While Kumuwa is correct regarding the proper test and the confusion apparent in the jurisprudence, it remains problematic as it looks to section 42(2)(e) of the Constitution for justification. Here, as correctly found in Uche and Kumwembe, the decision was based on a questionable interpretation of section 42(2)(e) which, according to the court, provides to sentenced prisoners a right to be released from detention with or without bail unless the interests of justice dictate otherwise. The problem the Court had to grapple with here is that section 42(2), on the face of it, does not apply to convicted, let alone sentenced, persons since it relates to persons accused of the ‘alleged commission of an offence’.
Sentenced prisoners only have a right to challenge the lawfulness of their detention and to be released if such detention is unlawful pursuant to sections 42(1)(e) and (f). Obviously, these rights are not the subject of an application for release pending appeal, if only for the reason that such an application is not a challenge to the lawfulness of detention within the meaning of section 42(1)(e). Holding otherwise would mean that the release of an applicant pending appeal is done pursuant to section 42(1)(f). This cannot be the case, since the decision to release is made before any finding on the substantive appeal and, thus, it cannot be said that such release is granted on the basis that the ‘detention is unlawful’ as dictated by section 42(1)(f). In any case, in view of the presumption in favour of a trial court’s sentence until it is set aside, the detention remains lawful until the appeal rules otherwise. Moreover, it would be difficult to explain the granting of release where the appeal is only against conviction and not against sentence. This understanding of sections 42(1)(f) and 42(2)(e) is supported by scholars such as Chirwa. He states that the former provision ‘will normally apply when there is no legal basis for an arrest or detention’, while the latter covers scenarios ‘where the state has good reasons for the arrest but fails to justify a request for further detention’ prompting the court to ‘consider whether the arrested person should be released with or without bail’.77
The problem is that, unlike the rights under sub-sections 42(2)(a) to (e) which are applicable before conviction, section 42(2)(f), while obviously falling under section 42(2), provides an ‘accused’ with the right to a fair trial which includes rights that are only applicable to convicted persons. This includes the right to be sentenced within a reasonable time after conviction and the right not to be sentenced to a more severe punishment than that applicable at the time the offence was committed. It becomes clear from the wording of section 42(2)(f) that one remains an ‘accused’ person until the appellate process has been exhausted. This contradicts the general flow of section 42(2). Justice Mwaungulu basically reasoned that, since some rights under section 42(2)(f) apply to convicted persons, the ‘accused’ in section 42(2) includes convicted and sentenced persons. This is true. However, the Court runs into problems when it holds that section 42(2)(e), which it says is ‘unfettered and unqualified in any respect’ as it does not use words like ‘before or after conviction’, also applies to convicted and sentenced offenders. This is inconsistent with the words ‘alleged commission of an offence’ in section 42(2). A convicted person is no longer an ‘accused’ in that sense. Therefore, on this score Uche and Kumwembe correctly faulted Kumuwa.
The interpretation of section 42 advanced in Kumuwa would actually lead to absurd results if only for the reason that convicted and sentenced persons would have rights under section 42(2) which they logically cannot have. For instance, sentenced prisoners cannot have the right to be segregated from convicted persons and to be presumed innocent until proven guilty under sections 42(2)(d) and 42(2)(f)(iii) respectively. The only plausible explanation to overcome such an outcome is that the definition of an ‘accused’ does not remain static throughout section 42(2). The problem here lies in drafting: Section 42(2)(f) should not have been included as a sub-paragraph to section 42(2), but as a subsection to section 42 itself. This would mean that section 42 would have three categories of rights, and the meaning of ‘accused’ would then have to be given contextual interpretation.78
A more fundamental problem with Kumuwa’s reliance on section 42 as the foundation for the interests of justice test is a procedural one emanating from the right to be released pending trial. The wording of section 42(2)(e) is such that the burden is on the state to show that the interests of justice are not in favour of release. In other words, there is a presumption in favour of release primarily premised on the presumption of innocence. Certainly, this cannot be the case with convicted or sentenced prisoners. If there is a right to release pending appeal, then it is for the state to show why release should not be granted to an appellant. Obviously, the reason will always be that the accused is serving a valid sentence of imprisonment. The onus, therefore, must be on the applicant to show why he should be released from custody in the face of a valid sentence of imprisonment.79 This would be inconsistent with the wording of section 42(2)(e) and case law that has emerged from it. Therefore, the right to be released under this provision cannot be the basis for a right to be released pending appeal.
However, it is not adequate to simply discard, as Uche and Kumwembe do, Kumuwa on the basis that section 42(2) rights are only applicable to suspects or, to use the language of the Constitution, persons ‘arrested for, or accused of, the alleged commission of an offence’. The overlap in the identity of the persons in section 42 is not limited to subsections (1) and (2), but also continues within the subsections. Some rights in section 42(2) do not accrue to suspects. This is particularly true of section 42(2)(f), which provides ‘an accused person’ with the right to a fair trial which, as confirmed in Kafantayeni & Others v Attorney-General & Others,80 extends to the post-conviction stage. As noted in Uche, some rights in section 42(2), such as the right to appeal and review, apply to both suspects and convicts. Other rights in this section cannot logically be claimed by suspects but only by convicts. These include typical sentencing rights, such as the rights to be sentenced within a reasonable time and not to be sentenced to a more severe punishment than that applicable at the time the crime was committed. Holding that section 42(2) excludes convicted prisoners also does not sit well with certain rights in section 42(2)(g) conferred on children in the criminal justice system. Good examples of this clash are the rights of children not to be sentenced to whole life sentences and to only be imprisoned as a last resort and for the shortest time possible.81 Equally, certain section 42(2) rights can only be claimed by suspects and not by convicts. This category covers rights such as the right to be presumed innocent, to challenge evidence and to be tried in a language which is understood by the suspect. It then becomes evident that section 42(2) requires a fluid definition of an ‘accused person’.
It could be argued that these difficulties arise because of the fact that sections 42(2)(f) and (g) are listed as subsections to section 42 and not as separate sections. Furthermore, the argument could proceed that the rights in section 42(2)(a) to (e) can only be claimed by suspects. This cannot be true. The right not to be compelled to make an incriminating confession or admission applies to both suspects and convicts. It would be folly for a court to hold that a sentenced prisoner does not have this right simply because he is no longer a suspect or that there is no allegation against him.
What, then, is the constitutional basis for release pending appeal? It is beyond the scope of the article to fully explore this issue. However, suffice to say that such release can be rooted in the right to liberty82 and a plethora of other rights that may come into play based on the facts of each case. For instance, the rights to life and health could be critical should it be that the continued detention of an applicant poses a real risk to his life in view of his ill-health. Release pending appeal can also be founded on the right to appeal ‘which presents the possibility that a conviction may be overturned’,83 making it paramount that convicted persons should be released from detention pending appeal when so warranted.
For some or other reason, Kumuwa did not consider the question of onus. However, there is also no suggestion in the judgment that the onus in release pending appeal applications would remain on the prosecution as is the case in applications for release pending trial. This lack of clarity could be responsible for arguments by the applicants in both Uche and Kumwembe that Kumuwa removed the need for an accused to show exceptional circumstances for release pending appeal. Clearly, it is this argument that informed the courts’ subsequent analyses of Kumuwa. Justice Mwaungulu does correctly conclude, albeit not clearly, that the circumstances that must inform the interests of justice test need not necessarily be unusual and exceptional.84 This is important given the fact that the depiction of a factor as usual or unusual, exceptional or unexceptional, cannot be made in a vacuum; it is a factual inquiry that rests on the facts of each case.
Overall, Kumuwa does not purport to change the traditional position which puts the onus on an offender to persuade a court to release him pending appeal. Therefore, it would be simplistic to argue that Kumuwa suggests that release before and after conviction is the same in all respects, including onus. Preposterously, placing the onus on the state means that the state must show why an applicant should not be released pending appeal. It cannot be the case that in an application for release pending appeal, the default should be release as is the case with release pending trial. While, as Chirwa correctly notes, ‘the state has an obligation to justify any form of detention’,85 this duty falls away when, as is the case with a sentenced prisoner, the person concerned is lawfully being detained as punishment. Once sentenced to imprisonment, the onus is reversed onto the accused. While the test for release remains the interests of justice, the default position is also reversed in that now release is to be denied unless the interests of justice require otherwise. This position is consonant with pre-trial release jurisprudence in Malawi:86
Bail being as of right to an accused person it should not be upon the accused to press for it when he appears before a court; bail should readily be available to an accused as he appears before a court and it should be upon the state to first give reason to the satisfaction of the court that bail should not be available to the accused ...
Transcribing this position to the situation of a sentenced prisoner is legally untenable. The assertion that offenders have a right to be released pending appeal, as flows from section 42, innately gives rise to onus issues as it necessitates the onus to be on the party, in this case the state, seeking to limit this ‘right’. As Justice Katsala queries in Macholowe v Republic:87
Now, if an accused person has the right to be released from detention then why should he be required to prove exceptional circumstances in order to be released on bail? ... [W]hy should he be required to prove exceptional circumstances before he can enjoy his constitutional right? Indeed, why should his release be in the discretion of the court when it is his constitutional right to be so released?
The use of comparable case law in Kumuwa is not convincing. To start with, the foreign law resorted to had to give insight regarding whether the right to be released from detention in section 42 of the Malawian Constitution is capable of being applied to sentenced prisoners, since this was the constitutional provision the Court was attempting to interpret. Disappointingly, the judgment does not adequately engage with South African case law on the point.
Smith, the case cited in Kumuwa, (apparently mistakenly) cites the interim Constitution of South Africa, of which section 25(2)(d) detailed that ‘[e]very person arrested for the alleged commission of an offence shall, in addition to the rights which he or she has as a detained person, have the right ... to be released from detention with or without bail, unless the interests of justice require otherwise’. This provision has been replaced by section 35(1)(f) of the final Constitution, which provides that ‘[e]veryone who is arrested for allegedly committing an offence has the right ... to be released from detention if the interests of justice permit, subject to reasonable conditions’.
Section 35(1)(f) removes any ambiguities on the scope of the right to be released from detention. This renders it less complex than the Malawian provision, which has accused and sentenced persons clumped together on this score of rights. Section 35(1)(f) leaves no room for an interpretation that it extends to convicted, let alone sentenced, persons. It ascribes the right to be released only to those ‘arrested for allegedly committing an offence’. The Constitutional Court in S v Dlamini exclusively tied this provision to persons ‘arrested for allegedly having committed offences, and may for that reason be detained in custody’.88 The Court also made it clear that the import of the change in wording between sections 25(2)(d) and 35(1)(f) was profound. As Justice Kriegler succinctly explained, while the former ‘contemplated release unless adverse factors tilted the scale against release’,89 the latter ‘require[d] something positive to permit release’ since ‘a person’s constitutional right to release from custody is now dependent on a finding that the interests of justice permit it’.90 Put differently, section 35(1)(f) changed the default position in that, while ‘previously ... an arrestee was entitled to be released’, the law now requires that ‘unless there is sufficient material to establish that the interests of justice do permit the detainee’s release, his or her detention continues’.91 Consequently, in pre-trial release applications, the onus is on an accused to persuade a court to grant him release. This is the exact opposite of the position in Malawi where, as noted earlier, the default is release. Therefore, it cannot be said that the South African law on bail is the same as Malawian law. Moreover, for certain categories of serious offenders, South African law requires that they may only be released pending trial on proof of ‘exceptional circumstances’.92 No lesser standard is expected from such an offender when seeking release pending appeal.93 The right to be released from detention if the interests of justice so permit ‘of course clearly deals with the position before finalisation of a trial’.94
Subsequently, Smith was incorrect in holding that the right in section 25(2)(d) extended to applications for release pending appeal. Be that as it may, Smith may be instructive on the fact that the presumption of innocence no longer plays a role post-conviction. Poignantly, this then makes it difficult to argue that a sentenced prisoner has a right to be released from detention, a right largely premised on the presumption of innocence. Indeed, it has often been said that pre-trial detention should not be used as punishment because at that stage the presumption of innocence stands. The position of a sentenced prisoner is the exact opposite of this: His detention is punishment.
The granting of release pending appeal is inherently different from release pending trial. While it is untenable to argue that the former stems from the right to be released from detention, it is also tenuous to restrict release pending appeal to situations where an offender has proved two factors, namely, the prospect of success on appeal and the likelihood of the sentence expiring before the appeal is concluded. Hence, Kumuwa was right in holding that the test for release pending trial and appeal was the interests of justice. However, it is incorrect to say that the constitutional premise for release remains the same before and after trial. Therefore, the High Court in Uche and the MSCA in Kumwembe were correct in finding that Kumuwa had been wrongly decided on this point. However, both decisions are disappointing to the extent that they fail to address the underlying rationale in Kumuwa and whether the proper test for release pending appeal was the interests of justice. The gist of Kumuwa is that courts have been overly restrictive in their approach to release pending appeal decisions because they focus on pre-determined exceptional circumstances instead of looking at what the interests of justice require in a particular case. What is required is for courts to be more generous in deciding which factors should inform the interests of justice in a particular case, bearing in mind the conviction, sentence and the right to appeal.
Once it is conceded that Kumuwa erred in extending the right to be released from detention to sentenced prisoners, it becomes clear that, as far as the law on release pending appeal is concerned, Kumwembe in essence is no different from Kumuwa. Importantly, although using different terminology, the two cases ultimately arrive at the same conclusion, which highlights three points: (i) that the proper test remains the interests of justice depending on the circumstances of each case; (ii) that case law merely provides examples of circumstances that would justify release pending appeal; and, therefore, (iii) that there is no closed list of circumstances that should justify release pending appeal.
Despite the problematic premise on which Kumuwa based the interests of justice requirement, it correctly finds that the proper test is the interests of justice. In light of the above, the proper test for release pending appeal must be the interests of justice. This should be informed by the circumstances of the case. The court’s discretion should not be limited to a pre-defined category of circumstances, but should be dependent on the facts of each case. Appeals or reviews should not be rendered nugatory by a denial of release. If this approcah is endorsed, factors such as a lost record, ill-health and even pregnancy should count in favour of release pending appeal. Justice nyaKaunda Kamanga reminds us in Nanseta v Mapwelemwe & Another that an appellate court has an obligation to ‘see to it that the appeal, if successful, will not be rendered nugatory’.95 In the case of an appeal against a sentence of imprisonment, a successful appeal will be rendered nugatory if the courts do not carefully consider release pending appeal, since the appellant will certainly have suffered punishment despite his lodging an appeal. As held in Nomale, if by the time the appeal is heard the sentence or even a substantial part of it has been served, the whole process is rendered nugatory.96
The question remains as to what the interests of justice require regarding release pending appeal. Since, characteristically, the concept of ‘interests of justice’ is not static, there can be no one answer to what factors should justify release pending appeal as it will always be a factual enquiry turning on the facts of each case. However, if understood as a factor and not a test, then releasing offenders whose records have been lost will not be a problem as it cannot be in the interests of justice, however understood, that such an offender must be detained although he is unable to exercise his right to appeal. Importantly, courts can also take a cue from the factors listed in the Bail Guidelines Act mutatis mutandis.97 One may be tempted to argue that it is merely semantics to say that the test is the ‘interests of justice’ and not ‘unusual or exceptional circumstances’. However, as Justice Mwaungulu reminded us in 1997, release pending appeal ‘is about exceptional circumstances showing that [the interests of] justice would be served by releasing a convict’.
12. M Kufa ‘Cry the beloved bail: Exceptional circumstances in interlocutory proceedings in the magistrate’s court’ 31 October 2012 http://macgregor kufa.blogspot.co.za/2012/10/cry-beloved-bail-exceptional.html (accessed 9 April 2016).
14. Manuel & Others v Republic (Ruling)  MWHC 92; Chimbanga v Republic (Ruling)  MWHC 59; Willard v Republic  MWHC 621; Republic v Kumwembe & Others (Bail Pending Appeal)  MWHC 570; S v Mkandawire; In Re: Application for Bail under s 355(1) of the Criminal Procedure and Evidence Code  MWHC 160.
17. See bail pending appeal application in Mwawa v Republic Criminal Appeal 50 of 2006 3 (referring to the fact that the accused had been granted release pending appeal on account of his poor health).
19. Sumaili v R [1961-1963] 2 ALR Mal 552 (HC), where the accused may be released on bail where the record of the lower court’s record had been questioned, necessitating the submission of affidavits from the court itself. This was done independently of the likelihood of success.
23. Nathebe v Republic Miscellaneous Criminal Application 90 of 1997 (HC). This flows from the fact that domestic problems are considered irrelevant to sentencing: See Republic v Mutawo Confirmation Case 237 of 1999; Republic v Eneya Criminal Case 53 of 2000.
25. This factor was rejected outright in Suleman (n 10 above) and Osman v Republic Miscellaneous Criminal Appeal 65 of 2004. However, the suggestion in Nwangwu v Republic Criminal Appeal 104 of 2007 and Chitapa v Republic Miscellaneous Criminal Application 190 of 2001 is that ill health is exceptional unless there is adequate medical treatment in prison.
26. This factor was rejected outright in Kaliati v Republic Miscellaneous Criminal Application 236 of 2006; Denja v Republic Miscellaneous Criminal Application 191 of 2001; Mila (n 15 above); and Chitapa v Republic Miscellaneous Criminal Application 190 of 2001, but accepted in Mponda v Republic Criminal Appeal 11 of 2005.
27. This factor was rejected outright in Chikwewa (n 13 above) but accepted in Nomale v Republic Criminal Appeal 178 of 2008; Mkandawire v The State Miscellaneous Criminal Application 175 of 2008; and Mussa v Republic Criminal Cause 65 of 2008.
44. Kumuwa (n 2 above) 5-6, referring to offenders sentenced under secs 337 to 340 of the CPEC based on factors such as the lack of a previous conviction, youth, old age, home surroundings, and the health or mental condition of the accused.
82. See R v Farinacci (1993) 18 CRR (2d) 298, finding that, although the right not to be denied reasonable bail is rooted in the presumption of innocence which is substantially spent by the conviction, the appellate process comes with a sufficient residual liberty interest that warrants the protection of the right not to be arbitrarily detained and, to some degree, the presumption of innocence.
84 Kumuwa (n 2 above) 5-6, observing that in view of the considerations in secs 337 to 340 of the CPEC which could ground a decision to grant release pending appeal, ‘the test of unusual and exceptional circumstances is not accommodative; unless, of course, courts regard matters raised in [these provisions] as unusual or exceptional. The rendition would itself be unusual and exceptional. The corollary, which is the better rendering, would regard the factors in [these provisions] as, as it should be, usual circumstances. The problem with this version is that it means that the test for granting bail pending appeal would have to be formulated [so] as to include these usual, unusual and exceptional circumstances. Such a test would be very pregnant as to be not a test at all. The test of unusual and exceptional circumstances could cause problems for the dual powers that are created in sec 355(1) of the [CPEC]. The power to stay execution of sentence is a substantial power that cannot be limited to unusual and exceptional circumstances. Equally, there are bound to be cases of [a] usual and unexceptional nature where, pending an appeal, the court may grant bail or stay execution of a sentence. In Rep v Keke (2010) Confirmation Case 404, this court considered regarding gender in sentencing offenders. The principle enunciated in that case is that sentencers must be sensitive to that, in treating male and female offenders in the same way, in some cases, offends section 20(1) of the Constitution of Malawi. It should, therefore, be very competent and compellable, for example, for a sentencing court at first instance or on appeal, based on the nature and seriousness of the offence and sentence, to stay the sentence of a woman or release on bail a woman who, because, in the nature of things, the woman is with child in the womb, something that, but for modern science and technology, would not happen to a man. All these considerations weaken or compromise the “unusual and exceptional” test for stay of execution of sentence and granting of bail pending appeal.’
97. The exception of capital cases is interesting. The court justified it by stating that all death sentences are automatically stayed by operation of secs 325 and 326 of the CPEC. These provisions dictate that all death sentences must be considered for mercy by the President. This does not explain why a death row inmate cannot apply for release pending appeal of his sentence and conviction. After all, the President’s intervention can only come after the MSCA has confirmed the sentence.