Africa, Prisons and COVID-19

Abstract Africa’s prisons are a long-standing concern for rights defenders given the prevalence of rights abuses, overcrowding, poor conditions of detention and the extent to which the criminal justice system is used to target the poor. The paper surveys 24 southern and east African countries within the context of COVID-19. Between 5 March and 15 April 2020 COVID-19 had spread to 23 southern and east African countries, except Lesotho. The overwhelming majority of these countries imposed general restrictions on their populations from March 2020 and nearly all restricted visits to prisons to prevent the spread of the coronavirus. The pandemic and government responses demonstrated the importance of reliable and up to date data on the prison population, and any confined population, as it became evident that such information is sorely lacking. The World Health Organization recommended the release of prisoners to ease congestion, a step supported by the UN Subcommittee on Prevention of Torture. However, the lack of data and the particular African context pose some questions about the desirability of such a move. The curtailment of prison visits by external persons also did away with independent oversight even in states parties to the Optional Protocol to the Convention against Torture (OPCAT). In the case of South Africa, prison monitors were not listed in the ensuing legislation as part of essential services and thus were excluded from access to prisons. In the case of Mozambique, it was funding being placed on hold by the donor community that prevented the Human Rights Commission from visiting prisons. The COVID-19 pandemic has highlighted long-standing systemic problems in Africa’s prisons. Yet African states have remained remarkably reluctant to engage in prison reform, despite the fact that poorly managed prisons pose a significant threat to general public health care.


Introduction
Africa's prisons are a long-standing concern of human rights defenders. Overcrowding, poor nutrition, illness, disease, and rights violations, to name a few concerns, characterize * The author is Associate Professor at the Dullah Omar Institute, University of the Western Cape, South Africa. He has been involved in criminal justice reform since 1992; his current focus is on the prevention and combating of torture and ill-treatment of prisoners and detainees.
Africa's prisons. The possibility that a highly infectious virus for which there is no vaccine becomes prevalent in Africa's prisons is a scenario almost too disastrous to contemplate. Many are familiar with the devastation that AIDS brought to Africa's prisons and with underlying health conditions such as HIV/AIDS, tuberculosis (TB) and general poor nutrition, COVID-19 does conjure up images of an unprecedented prisons health crisis, as both the infection rate (the proportion of people infected) and the infection fatality rate (the proportion of infected persons who die) may be higher in prisons than that which may occur among the general population outside prison. This essay reviews the situation regarding prisons and COVID-19 in southern and east Africa. There are many gaps in the information, which does point to long-standing issues on data reliability and accessibility. Nonetheless an overview of prison numbers is provided, as well as some general comments on conditions of detention. This followed by a short description of COVID-19 in Africa and how different states have responded, imposing more or less severe restrictive measures.

Overview of the numbers
Southern and east Africa consist of 24 internationally recognized countries 1 and have a combined prison population of some 655,000 people. 2 The average national prison population size for these countries is some 30,000, with the average prison holding 533 prisoners. There are, however, significant outliers with island states like the Comoros having 225 prisoners in three prisons and, at the other end of the spectrum, South Africa with some 235 prisons holding close on 160,000 prisoners at an average size of 657 prisoners per facility. A further important variable is the proportion of prisoners who are awaiting trial. This is important for two main reasons. Firstly, it indicates the efficiency and effectiveness of the criminal justice system, with a low percentage indicating that there is good reason for pretrial detention and that individuals do not remain in pretrial detention for an excessive time. Secondly, a high percentage of pretrial detainees not only reflects inefficiency and ineffectiveness, but also results in more contact between the prison population and the outside world, because of remand prisoners' appearances in court and visits to them in prison, and this contact is relevant to the COVID-19 discussion. In the 24 southern and east African states surveyed, the highest proportion of pretrial detainees is found in Madagascar at 57 per cent and the lowest in Rwanda at eight per cent. The average for all 24 countries is 31 per cent, meaning that nearly one-third of bed spaces in prisons are occupied by people presumed to be innocent. The last important variable to note is the level of occupancy. In Uganda and Zambia prisons are occupied at 319 per cent and 303 per cent respectively, meaning three people are using the space originally designed for one person.
Colonial-era laws as well as colonial-era modalities of policing have also shaped to some extent the profile of that segment of the population coming into (regular) contact with the criminal justice system. It is in particular in anglophone states of southern and east Africa that a myriad of laws dating back to the previous century give the police wide discretion in arresting people for poorly defined offences such as loitering or being a rogue and vagabond (Muntingh 2015;Muntingh and Petersen 2015). These petty and outdated offences remain an affront to human dignity and perceptions of fairness, and need to be removed from statutes (ACHPR, Principles on the Decriminalisation of Petty Offences in Africa). Nonetheless, their continuous enforcement plays an important role in determining who is imprisoned.

Conditions of detention
While there is much variation in the conditions of imprisonment in these countries, it can generally be concluded that there is significant room for improvement and, in some instances, prison systems are not able to meet even the most basic standards of humane detention with reference to accommodation, nutrition and health care. In some states prisoners are provided with only one meal per day of the most basic nature, resulting in malnutrition; illness and infectious diseases are common in the prison population; and prisoners are forced to sleep in cells filled to two or three times their capacity with the most rudimentary toilet facilities, and with limited capacity to maintain and keep ablution facilities sanitary. These conditions worsen health outcomes. Poor conditions of detention and structural problems in African criminal justice systems are well documented and a number of efforts have been made at international and regional levels to enable, encourage and support prison and criminal justice reform. The African Commission on Human and Peoples' Rights (ACHPR) not only appointed two special mechanisms 3 on the issue, but also adopted four soft law instruments to support reform. 4 Despite these efforts as well as general efforts to improve conditions of detention, the situation has not improved markedly. There are a few countries in east and southern Africa that have been able to effectively maintain and replace (if not expand) prison infrastructure, such as South Africa and Botswana. But in most instances, the same infrastructure built by Britain, Portugal and France as colonial powers more than 50 years ago remains in use and it is simply not possible to meet the minimum standards of humane detention (UN Standard Minimum Rules for the Treatment of Prisoners 2015). In many instances sentenced and unsentenced prisoners are not segregated, as is required by Rule 11(b) of the 2015 UN Standard Minimum Rules, and this has implications for controlling infectious diseases.
Not only have post-independence governments largely failed to invest in criminal justice and prison reform; they have also failed to see prisons from a public health perspective and give recognition to the inescapable fact that prisons are vectors for disease.

COVID-19 comes to Africa
The first COVID-19 case in Africa was reported from Egypt (BBC News 2020) and the second from Algeria (WHO Africa 2020). In southern and east Africa, the first case was reported in South Africa on 5 March 2020 and the first prison system case (an official) was reported also in South Africa on 6 April (GroundUp 2020; Graaff-Reinet Advertiser 2020). COVID-19 came to Africa through various entry points and most likely travel hubs like Johannesburg and Nairobi, from where it moved quickly to even the island states of the region, such as Seychelles (14 March

Lack of data and releases
The COVID-19 pandemic and subsequent government responses again demonstrate the importance of reliable and up to date data on the prison population, and any confined population (such as police and immigration detainees). This not only concerns data on the health of prisoners, as well as more general data concerning the profile of prisoners, the reasons for their custody and so forth, but above all, if this information is available in the public domain those with an interest in prisoners' rights can make informed contributions to governments and other stakeholders. When the World Health Organization (WHO) recommended the release of prisoners to address overcrowding (WHO 2020) and this call was further supported by the UN Subcommittee on Prevention of Torture (SPT) in its public advice to National Preventive Mechanisms (NPMs) (UN SPT 2020: para 9(b)), it seemed in principle like an admirable idea, but reality demanded a more cautious approach, in Africa at least. Firstly, cognizance should be had for public safety and that it cannot merely be a case of opening the prison gates indiscriminately, not least because the person concerned may currently be infected. This means that there need to be clear criteria for eligibility, a confirmed and contactable address, and presumably consent from the household in which the person is intending to reside. Moreover, should only deserving pretrial detainees be released or are sentenced prisoners also eligible prior to the expiry of their sentence? Following from this, it must also be asked: how many people need to be released in order to have a protective impact for those that remain behind? Secondly, in some instances, especially when an infection has not been recorded at a particular prison, it may in fact be a better option not to release prisoners since their basic needs are at least met in prison (this is assuming that basic nutritional and health requirements are complied with). Thirdly, assuming that the preceding requirements are satisfied, the question becomes: where do released prisoners go to? They may not be welcome at their normal place of residence, or they may indeed have no place to go to, or their return may place additional socioeconomic strain on an already marginal household. With large-scale loss of income in the formal and informal sectors, and the absence of state-provided socioeconomic support, the large-scale release of prisoners may require a re-think in the African context.
Regardless of this, a number of southern and east African countries announced the release of prisoners to 'ease congestion', but the intended impact would be symbolic rather than real. Seven countries announced significant releases with the highest number in South Africa (19,000), Ethiopia (5,600) and Mozambique (5,032) (Republic of South Africa, The Presidency 2020; Ethiopian Monitor 2020a; Daily Nation 2020) and the lowest in Botswana (149). In the case of South Africa, the releases would constitute 12 per cent of the total prison population. However, in the case of Ethiopia the prison system capacity is not known and the impact is thus uncertain, but the 5,600 releases constitute less than five per cent of the prison population. The 5,032 released in Mozambique moved the occupancy level from 211 per cent to 186 per cent (World Prison Briefs 2020). With the exception of Botswana, the releases effected did not move prison occupancy below the specified levels, and overcrowding remains present as ever. The releases should also be seen against the background of normal releases when prisoners are released conditionally (for example, parole) or on expiry of sentence. These figures are not in the public domain and it is therefore not known if the releases will have any meaningful and sustained impact on prison congestion.

High turnover of people
Much as we would like to think that prisons are isolated from society, they are not. Family and friends visiting prisoners are in many ways the lifeblood of the prison, bringing not only human interaction and contact with the outside world but also resources such as cash, food, bedding, toiletries and so forth. It is predominantly the poor who find themselves in prisons, but since the state cannot provide all the resources, prisoners are dependent on outside support (when they have such support) to make life in prison more bearable. In the end it is the poor who are subsidizing imprisonment in Africa, picking up the tab where the state falls short in providing the basics to prisoners (Muntingh and Redpath 2018). Admittedly, visitors also bring contraband fuelling the carceral economy-which is often crucial for some prisoners to 'buy' their safety. Contact with the outside world is vital for both physical and mental well-being. It is, however, not only prisoners' visitors who come to a prison. The staff working there as well as other officials also come to prisons on a daily basis. Prisoners awaiting trial and sentenced prisoners with further charges also have to appear in court. In short, the gates of a prison are busy. This has obvious implications for virus transmission.

Restrictions on access to prisons
Of the 24 countries surveyed in southern and east Africa, only two could be confirmed as not having declared a state of disaster or emergency or lockdown: Malawi and Tanzania (New Frame 2020). In the case of Malawi, the government did declare a lockdown, but this was blocked in the High Court (News24 2020). Eighteen of the 24 states surveyed declared some form of a national restriction, such as a state of disaster or a state of emergency. This was then to a greater or lesser degree supported by a lockdown or curfew (terminology differs from one state to another). In some instances, this was announced for a very limited duration (such as 21 days), but frequently extended. In ten states access to prisons by external visitors was prohibited (Independent Online 2020b; All Africa 2020a; Independent Online 2020c; Ethiopian Monitor 2020b; ICRC 2020a; Léger 2020; All Africa 2020b; ICRC 2020b) and in a further 11 states this could not be confirmed. South Sudan imposed light restrictions (that is, sign a register and wash hands) (UNDP South Sudan 2020), and Zambia prohibited visits but visitors could leave items for prisoners, such as food or clothing (Global Press Journal 2020). While it must be appreciated that governments have an obligation to protect their citizens against public health threats, it must also be noted that the states surveyed have by and large shown a particular enthusiasm to restrict civil and political liberties and, further, to limit access to prisons.

How did oversight agencies perform?
The guidelines from WHO made it clear that the 'COVID-19 outbreak must not be used as a justification for objecting to external inspection of prisons and other places of detention by independent international or national bodies whose mandate is to prevent torture and other cruel, inhuman or degrading treatment or punishment' (WHO 2020). Of the 24 countries surveyed, six have ratified the Optional Protocol to the Convention against Torture (OPCAT) 5 and two have signed it but not ratified. 6 In the case of South Africa, the Judicial Inspectorate for Correctional Services (JICS) provides regular visits to prisons through its inspectors as well as community-based Independent Correctional Centre Visitors (ICCV). However, the State of Disaster declared on 15 March and subsequent regulations (26 March) closed all prisons to civilian visits (Disaster Management Act Regulations) and did not include the JICS inspectors and ICCVs in the definition of essential services; thus they were not permitted to conduct prison visits. The net result was that not only were visits by friends and families to prisoners stopped, but so were the visits by ICCVs mandated under the Correctional Services Act. Moreover, this also prevented visits under the OPCAT, to which South Africa is a party, since the JICS forms part of the National Preventive Mechanism (NPM), the South African Human Rights Commission (SAHRC), as designated by South Africa in June 2019. While officials of the Department of Correctional Services (for example, prison warders) are exempted from the restrictions of the lockdown that commenced on 26 March, the JICS inspectors and ICCVs are appointed under a different provision of the Correctional Services Act resulting in them not being exempted from the lockdown. It then appears that a major victim of COVID-19 was transparency of the prison system. South Africa's neighbour, Mozambique, declared a state of emergency on 30 March (All Africa 2020c), but technically also a state of disaster in terms of a threat to public health or a natural disaster, and so which should be distinguished from martial law. The impact of restrictions imposed is less severe, and essentially, gatherings of more than 20 people are prohibited, including in prisons. Moreover, no restrictions, save for family members, are placed on who visits prisons and lawyers and the NPM is free to conduct visits and even training programmes are continuing as long as groups remain under the 20-person ceiling (T. Lorizzo, Director: REFORMAR, based in Maputo, Mozambique, interview with the author, 6 May 2020). Visits by the NPM are, however, not taking place since the donors 5 Mozambique, South Africa, Madagascar, Mauritius, Rwanda and South Sudan. 6 Angola and Zambia. that would have supported the NPM's visits to places of detention (and planning was reportedly well under way) placed all funding on hold due to the COVID-19 pandemic.

Conclusion
COVID-19 has again highlighted the long-standing problems in Africa's prisons, and these were noted above. As unpopular as this may sound, Africa needs new prisons, not necessarily more prisons, because the basic infrastructure is so old and dilapidated that it is simply no longer possible to meet the minimum requirements of humane detention, let alone to have the capacity to deal with a health crisis. It is necessary to look beyond COVID-19 and pose questions about long-term solutions.
People should also not find themselves unnecessarily in prison and that means that reforms in law, policy and practice are needed, as has been advocated by many, including the African Commission on Human and Peoples' Rights. Pretrial detention is very often the default position without any real exploration of conditional release mechanisms such as warning and bail. Monetary bail is frequently set at amounts that are simply not affordable (such as equivalent to a year's income) that can only be regarded as punishing people for being poor.
The statutes of many African states still contain colonial-era offences (such as rogue and vagabond) and in a more recent development, the creation of new petty offences, frequently under municipal by-laws, for which people are arrested and end up in custody. Placing people in custody for petty offences is not only disproportionate, it places their health at risk and holds significant adverse socioeconomic consequences. These offences need to be decriminalized and/or declassified.
Prison overcrowding can to a certain measure be addressed by using non-custodial sentencing options, such as community service, or releasing prisoners prior to the expiry of sentence (for example, on parole). There are only a few African states where these measures are used on any significant scale.
Public health threats such as COVID-19 or tuberculosis (TB) will always have particular significance in the prison environment, but also pose a particular threat to public health when prisons are poorly managed and resourced. In the end, the old dictum stands: Good prison health is good public health.

Funding
This work was supported by the Open Society Foundation (Human Rights Initiative) and the Sigrid Rausing Trust.