Epidemics and the Military: Responding to COVID-19 in Uganda

The UN Security Council's response to Ebola in 2014 legitimised militarised responses. It also influenced responses to COVID-19 in some African countries. Yet, little is known about the day-to-day impacts for ordinary citizens of mobilising armies for epidemic control. Drawing on 18 months ethnographic research, this article analyses militarised responses to COVID-19 during, and following, two lockdowns at contrasting sites in Uganda: a small town in Pakwach district and a village in Kasese district. Both field sites lie close to the border of the Democratic Republic of Congo. Although the practice of health security varied between sites, the militarised response had more impact than the disease in these two places. The armed forces scaled back movement from urban conurbations to rural and peri-urban areas; while simultaneously enabling locally based official public authorities to use the proclaimed priorities of President Museveni's government to enhance their position and power. This led to a situation whereby inhabitants created new modes of mutuality to resist or subvert the regulations being enforced, including the establishment of new forms of cross-border movement. These findings problematise the widely held view that Uganda's response to COVID-19 was successful. Overall, it is argued that the on-going securitisation of global health has helped to create the political space to militarise the response. While this has had unknown effects on the prevalence of COVID-19, it has entrenched unaccountable modes of public authority and created a heightened sense of insecurity on the ground. The tendency to condone the violent practice of militarised public health programmes by international and national actors reflects a broader shift in the acceptance of more authoritarian forms of governance.


Introduction
COVID-19 unfolded in different socio-political contexts, with varied consequences. In Europe, it was initially assumed that transmission patterns would correspond to outbreaks of influenza. Meanwhile, in some parts of Africa, responses were shaped more by previous experiences of Ebola. Crucially, that included the deployment of armed forces with effects that have been largely overlooked.
This article focuses on Uganda, a country that has had experience with Ebola since 2000. Particular attention is given to the way the military intervened during and following COVID-19 lockdowns between March 2020 and December 2021. We ask: how have recent trends in disease control by the military emerged from global and national processes? What are the day-to-day impacts of sustained military engagement for ordinary citizens? The term 'military' is used to refer to armed personnel at least nominally acting as an aspect of formal governance, including soldiers, police and militia.
The article begins by outlining the historical issues shaping the interconnections between health, security and the military. This is followed by a discussion of methods, and national responses to  To elucidate the day-to-day practice of health security, two case studies are presented from a small town in Pakwach district, north-western Uganda and a village in Kasese district, western Uganda. We show that forms of public authority have been enhanced that reinforce unaccountable governance, link disease control with violent abuse, and are resisted in ways that are unlikely to be associated with improved health outcomes.

Health, security, and the military in historical context
Debates about the role soldiers and enforcement should have in public health have a long history; and militaries have been an aspect of medical humanitarianism since its inception (Allen, 2018). The Red Cross attempted to maintain 'neutrality' while working alongside, and even within, armed forces, and it is no coincidence that several international health institutions emerged out of wars. That includes the World Health Organization whose constitution states: "[t]he health of all peoples is fundamental to the attainment of peace and security …" (World Health Organization, 1948;July 22). This phrasing evokes the phrase used in Chapter VII of the United Nations Charter that legitimises "action by air, sea, or land forces as may be necessary …" Nevertheless, there have been important changes in militarised trajectories of health care since the end of the Cold War, reflected in a burgeoning literature (e.g., Lakoff and Collier, 2008;Rushton, 2011Rushton, , 2019Lancet, 2014;Michaud et al., 2019;Wenham, 2019;Gibson-Fall 2021). Two interrelated trends since the mid-1990s have been particularly significant, and the way they underpinned responses to the West African Ebola outbreak has had enduring consequences.

Humanitarian assistance and intervention
First, it has become increasingly difficult to distinguish between humanitarian assistance to those in acute need and humanitarian intervention to forcibly prevent atrocities. This was evident in, for example, Somalia, Kosovo, and Afghanistan, and became a matter of serious concern for some medical non-governmental organisations. That was notably the case for Médicines Sans Frontièrs (MSF), which vigorously opposed military involvement in health programmes, even though the option of armed intervention, based on 'a right to interfere' was actively promoted by one of the agency's founders (Allen and Styan, 2000). Also, by the end of the 1990s, there was widespread recognition of failures in humanitarian responses to particular crises, especially those in Rwanda and former Yugoslavia. This underpinned the emergence of Responsibility to Protect (R2P), a concept that was eventually adopted by UN member states at the World Summit in 2005. MSF vociferously opposed the implications of R2P because it effectively justified militarised interventions (Weissman, 2010). Nonetheless, despite R2P having mixed results in Libya and Syria (Ainley, 2018), the idea of armed deployments for the purpose of promoting well-being has become a more established possibility.

Globalisation and health threats
Second, several analysts drew attention to the potentially adverse consequences of globalisation, including the potential health implications. Laurie Garrett's The Coming Garrett, 1994, became a best-selling book. She, and other analysts, directed attention to dangers associated with air travel, integration of food markets and environmental changes for the spread of new diseases (e.g., Fauci, 2001;Morens et al., 2004;Rodier et al., 2000). Their warnings were underlined by the experience of HIV/AIDS in the 1980s, SARS in 2002-2005, HINI in 2009, MERS in 2012, Ebola in 2014 influenzas. There were also concerns about biological terrorism after Anthrax attacks in 2001 (Heymann, 2003;Fidler and Gostin, 2006) and increasing alarm about antimicrobial resistance (Heymann and Rodier, 2001). A host of publications emphasised the need for better strategies to monitor and contain infections, including a World Health Report (World Health Organization, 2007). Echoing the WHO constitution, the literature explicitly evoked links between health and security, but, in the context of upheavals during the 1990s and the discussions about R2P, the policy implications of the word 'security' shifted. Previously, 'health and security' had mainly focused on the health of armed forces and how health and health care were influenced by conflict (McInnes 2015); or foregrounded epidemics that occurred centuries ago (e.g., Wenham, 2019), or, especially in the context of HIV/AIDS, the balancing of risks and insurance (Elbe, 2008). Now, health security was being connected more overtly with militarised security, direct threats to peace, and even to the possibility of pathogens being weaponised.

Security Council Resolution 2177
A key event, connecting these two trends, occurred when the United Nations Security Council passed Resolution 2177 in September 2014 (United Nations Security Council, 2014:1). The Resolution described the Ebola outbreak in West Africa as 'a threat to international peace and security' and legitimised enforcement action. Some scholars expressed disquiet at the time (de Waal, 2014;Abramowitz et al., 2014) and several assessments of the outcome were critical (DuBois et al., 2015;McKay and Parker, 2018). However, others were at least initially supportive, including MSF (Allen, 2018).
Whatever the merits of the decision, it reflected a fundamental shift in international attitudes from those reflected in Security Council Resolution 1308, dealing with HIV/AIDS (UN Security Council, 2000). That resolution, adopted fourteen years previously, was described in relation to the Security Council's responsibility to maintain international peace and security, but it did not legitimise enforcement action. Instead, it encouraged United Nations Member States to acknowledge the problem of HIV/AIDS, especially in Africa, and to consider voluntary HIV testing for troops involved in peacekeeping. The contrast in tone is striking.

Militarisation of health care
These changes in international perspectives are important, although it should not be overlooked that militarised health care has different histories in different places. In former European African colonies, disease control involved policing and sometimes armed policing (Vaughan, 1991). Acts such as mass forced displacement were enforced on a large scale to contain infectious diseases in some regions, and violations of even relatively minor public health decrees could be met with beatings. In such contexts, recent recourse to enforcement of curfews, quarantine, and/or restrictions of movement during the COVID-19 pandemic could legitimately be seen as more of the same, rather than a new kind of public health. The use of militaries in, for example, Uganda (Akello, 2020;Parker et al., 2020), South Africa (Manderson and Levine, 2021) and Nigeria (Iweze, 2020), was thus rather less surprising than Italy (Reuters, 2020a), Spain (Reuters, 2020b) and the UK (Guardian, 2020).
This does not mean that military deployments are necessarily less troubling among African populations. On the contrary, it is possible that previous experiences with armed forces as forms of local public authority might create situations whereby soldiers become involved in extracting rents to enhance their livelihoods, rather than containing disease outbreaks. Moreover, where military logic becomes entrenched, information flows may be restricted, and objectives involving 'collateral damage' normalised. This is most likely in contexts where democratic institutions are weak or partial, and where those at the receiving end of public health control measures have little voice. It makes it important to assess what happens in practice. However, there is a paucity of such research.
Some scholarship has highlighted the way in which militarised metaphors have been mobilised to help justify enforcement measures and the infringement of civil liberties (e.g., De Waal, 2021). Other scholars have highlighted acts of resistance, particularly during the 2013-2016 West African Ebola outbreak. (e.g., Benton, 2017;Wilkinson and Fairhead, 2017;Parker et al., 2019a). Some of this research is based on ethnographic fieldwork carried out soon after a specific event, but most of it analyses accounts provided by journalists and other kinds of social media. Generally missing is any reflection on the day-to-day impacts of sustained military engagement for ordinary citizens, especially in rural areas. The ethnographic findings presented in this article respond to this gap. Data are analysed with a public authority lens, with public authority defined as 'any kind of authority beyond the immediate family which commands a degree of consent' (CPAID, 2018:8). As indicated elsewhere (Parker et al., 2019a;, this lens focuses on the relationship between formal, informal, parallel and hybrid authorities and helps to reveal socio-political dynamics that might otherwise remain obscure. It enables an exploration of the (dis) connections between global framings of health security and the national and local politics influencing the diverse ways in which public health security is enacted on the ground.

Methods
Research is based on multi-sited, historically informed ethnographic fieldwork. It began in July 2019 and pivoted from a broad focus on epidemic preparedness and response in contexts where people live with different kinds of hazards (such as floods, drought, food insecurity) to a specific focus on COVID-19 in March 2020 (see MacGregor et al. 2021 andLeach et al., 2022  Thirty-two open-ended, unstructured interviews were also carried out by MP and GA with national actors involved in the work of the national task force. These interviews provided an opportunity to discuss specific issues arising during the zoom calls (e.g., diagnostic testing for , and to assess the way in which national and international discourses on public health security were shaping the design and implementation of national policy seeking to control COVID-19.
Research at a district level involved GA, BO and MB doing 10 interviews with staff contributing to the COVID-19 district task force meetings in Kasese district; and PK doing 5 interviews in Pakwach district. These open-ended, unstructured interviews provided an opportunity to explore the way in which members of the armed forces became embedded in the district task forces, and to discuss the various strategies they deployed to implement national policy on the prevention and control of COVID-19.
Village and town-based fieldwork involved participant observation, including open-ended, unstructured interviews and semi-structured interviews with a range of people including local council chairpersons, health assistants, schoolteachers, farmers, fisherfolk, members of village health teams, herbalists, rainmakers, and motorbike-taxi riders. It began in July 2019 in Omusya village (a pseudonym), Kasese district; and the small town of Panyimur, Pakwach district in April 2020 (see further on for details about these places). Fieldwork at the latter site built on research carried out between 2005 and 2012 on neglected tropical diseases (e.g. Parker et al., 2008;Parker and Allen, 2011). Research officers (MB and PK) were permanently located in Omusya and Panyimur respectively; and another research officer (BEO) was resident in Panyimur for the month of April 2020. He subsequently made five 7-10 day follow-up visits to the town. Fieldnotes from both sites were shared by email (usually weekly), and feedback was provided by a combination of written and oral communication. This enabled the day-to-day impacts of the militarised response to COVID-19 to be followed up in a continuous, dynamic and iterative way. When global and national travel restrictions eased, shorter fieldtrips were made to the village of Omusya by GA and BO, and Panyimur by MP, BO and TA.
The next section draws on findings from our national level research as well as accounts in social media and the wider academic literature to outline the national response to COVID-19. Attention is given to the historical and political context in which COVID-19 unfolded. The section provides important contextual information for ethnographic findings detailing the day-to-day practice of health security in Pakwach and Kasese districts.

National responses to COVID-19
In Uganda, as elsewhere in Africa, there is a long history of enforcement action in containing epidemics, including the movement of populations to contain trypanosomiasis under British colonial rule (Lyons, 1992;Vaughan, 1991), and the implementation of behavioural measures to prevent HIV transmission during the mid-late 1980s and 1990s (Allen, 2006). In 2000/2001, the Ugandan Peoples' Defence Force (UPDF) was mobilised to identify cases of Ebola during an outbreak in Gulu, northern Uganda (Kinsman, 2012); and the armed forces were regularly deployed in western Uganda to restrict cross-border movement and prevent the spread of Ebola from neighbouring DRC (Akello and Parker, 2021).
With respect to COVID-19, WHO declared the virus to be a Public Health Emergency of International Concern on January 11th , 2020 and a pandemic on March 11th , 2020. There were no reported cases of COVID-19 in Uganda during this time. The Ministry of Health's Public Health Emergency Operation Centre (PHEOC) was busy winding down operations which had been put in place to respond to anticipated outbreaks of Ebola in Kasese district (which borders DRC), while simultaneously trying to elicit support from international donors to respond to outbreaks of typhoid, cholera, yellow fever, malaria, TB, and anthrax in other parts of the country.
Nevertheless, President Museveni responded quickly to the WHO announcing a pandemic. Instead of deferring to the PHEOC, he militarised the response. This involved establishing a COVID-19 National Task Force and imposing a command-and-control system with operations running through the Office of the Prime Minister (Parker et al., 2020). A lockdown was swiftly announced and the UPDF enforced it, with additional support coming from paramilitary Local Defense Unit's (LDUs). The UPDF and LDU's had a visible presence on streets and highways. They were tasked with preventing movement across international borders (unless it involved cargo); checking that people had elicited the relevant permission letters to travel; ensuring the closure of markets, churches, mosques and other public places; providing deployments to ensure quarantine at designated centres was carried out appropriately; checking that burials, weddings and homestead gatherings were carried out in accordance with official Standard Operations Procedures. Curfews were strictly enforced.
Beyond the military, the COVID-19 National Task Force coordinated activities across complementary pillars. (Ministry of Health 2020). They were run by civil servants and high-ranking military officials, with advice being provided by WHO and donors such as USAID, CDC, DfID/FCDO and UNICEF. To disseminate and implement policies quickly across the country, COVID-19 task forces were also established at district and sub-county levels, and eventually at a village level. Resident district commissioners are political appointments, and they chaired the district taskforces, with district health officers mandated to take the minutes. Police commanders and internal security officers were also required to attend these meetings.
The speed with which President Museveni responded to the pandemic in March 2020 was welcomed (Kitara and Ikoona, 2020); and initially there was widespread support for the first lockdown, which was announced on March 18th , 2020. Metaphors used by the President to justify the lockdown included phrases such as "This is a war of the wanachi [people]. I'm here to lead the wanachi war …". He went on to say that anyone challenging the approach would be treated with "iron gloves deserving of enemies" (March 30th , 2020). This was not an empty threat. Two leading members of the opposition, Bobi Wine and Kizza Besigye, attempted to distribute food to people who had lost their source of income during lockdown. Museveni responded by saying: "I direct police to arrest people who will be distributing food to people. That is looking for cheap popularity … you will be charged with attempted murder" (Daily Monitor, 2020).
The political context in which COVID-19 unfolded explains the ferocity of Museveni's responses, and it gradually became apparent that the biosecurity measures put in place served his political interests well. Museveni came to power in 1986. In 2017, parliament agreed to remove the 75-year age cap for presidential candidates. He was 73 years old at the time, and the constitutional change enabled him to stand for reelection in January 2021. By March 2020, it was clear that he was going to face considerable opposition in some parts of the country. Imposing lockdown thus provided an opportunity to prevent crowds gathering during the election campaign, while also helping to inject fear and limit debate. Nevertheless, there were numerous occasions when violence and unrest spilled onto the streets. The army was reported to have responded brutally (e.g., Biryabarema, 2020).
President Museveni eased restrictions on June 2nd , 2020. This was partly because 590 cases and zero fatalities from COVID-19 had been reported (figures that were low compared to other nations), but also because of public outcry about the challenges of sustaining a living in lockdowns without state assistance. Although most public venues remained closed and political rallies were forbidden, some aspects of life returned to 'normal'. It was possible to move between districts within Uganda, and to buy and sell goods in markets. The official narrative shifted to one of blame and responsibility. This is epitomised in a disingenuous speech President Museveni made on national television: Missing from the speech, and others which followed (https://www.st atehouse.go.ug/media/speech), was any recognition of the serious shortage of 'staff, stuff, space and systems' (to use Farmer's well-known phrase (Farmer, 2014). Uganda has an estimated population of 45 million, but at the time of the first lockdown, the main teaching hospital in Kampala -Mulagoonly had capacity to provide critical care for 36 adults and 27 children at any one time. A further 35 beds were available at the Women's Hospital in Kampala, while the country's 117 Regional Referral Hospitals had capacity to provide intensive care for an average of 10 people per hospital at any one time. Testing for COVID-19 was also limited, with the Ugandan Viral Research Institute initially being the only place able to provide PCR testing (personal communication with Professor Pontiano Kaleebu, April 2020).
Colleagues on the National Task Force were acutely aware of the situation and the urgent need to source diagnostic equipment, medicines and protective clothing. Indeed, the limited resources available for the treatment and prevention of COVID-19 was indicative of a paucity of health care more broadly. At study sites in Pakwach and Kasese districts, for example, routine child immunisations are provided by unsalaried village health workers. Maternal and obstetric care typically involves travelling to a health centre (which can be 60 minutes walk away). Some diagnostic testing is intermittently available for a range of infections (including malaria, typhoid, HIV/AIDS and pneumonia) at these health centres. However, stock-outs are common, and the cost of treatment is often prohibitive. Seeking critical care for COVID-19 would have been much too expensive for the vast majority of study participants, even if it had been available.
It is also important to note that even though lockdown restrictions were eased for a year (June 2nd , 2020 to June 6th , 2021), schools remained closed and public gatherings were prohibited throughout this period. Opposition parties were thus unable to campaign, and this is likely to have contributed to Museveni's victory in January 2021, although electoral rigging is a possibility too. Shortly before the election, EU monitors departed stating that there was no indication that the election would be fair (Biryabarema, 2020). Their assessment has been corroborated by others (Abrahamsen and Bareebe, 2021).
Following the inauguration of the President in May, a second lockdown was announced on June 6th, 2021. This happened at a time when cumulatively a total of 52,935 cases and 383 deaths had been reported (http://coronavirus.jhu.edu/region/uganda), and infections from the Delta variant were rising fast across the country. Hospitals in Kampala, Entebbe, Gulu etc were full, and medical oxygen was in short supply. In contrast to the first lockdown, it was possible to move within a district and markets remained open. However, schools, places of worship and bars remained shut, and curfews were strictly enforced. The next section presents case studies from Pakwach and Kasese districts to elucidate the practice of health security.

Case study from Pakwach district
Fieldwork took place in a small town called Panyimur, on the northern shores of Lake Albert, a few miles from the international border with DRC. Most residents are Alur. Livelihoods are based on a portfolio of informal activities, with most people, including those formally employed as teachers, health workers and local politicians, involved in fishing and subsistence agriculture. Traders regularly travel to the town from DRC, South Sudan and parts of northern, western and southern Uganda to buy and sell fish, charcoal, petrol, timber and agricultural produce (Parker et al., 2008;Titeca and Flynn, 2014).
The first lockdown was characterised by extensive violence. Within hours of the announcement, the army arrived at the busy market. They fired live rounds of ammunition into the air and beat traders until they ran awayalthough few people knew that trading was prohibited. Over the coming weeks and months, the UPDF enforced lockdown measures with the support of the police, immigration officials and LDU's. These were reportedly enforced brutally, and a diverse group of people were affected. They included: young women running small kiosks in the trading centre, adults watching films with their families in their compounds, and motorcycle riders taking ill relatives to health centres in other parts of the district. The following account from a pastor was not untypical: "That fateful day I had a lot of work to do in the community. I and my prayer group did pastoral work in five different homesteads, and by the time we were finishing the last one, it was already late (past 7pm). So I hurriedly jumped on my motorcycle and left for home. But before I could reach home, the security forces stopped me and without saying a thing, they descended on me with sticks, I lost control and fell down … they beat me until I lost consciousness." Fisherfolk were also targeted, especially those returning with boats full of fish after curfew. As Westaway et al. (2007) and Parker et al. (2012) have pointed out, they have long been perceived as 'feckless' and 'reckless' by government officials, with their erratic lifestyles making them hard to govern. On the northern shores of Lake Albert, this has created enduring tensions. Officials, all too often, assume that they dodge taxes and pay no attention to restrictions on net sizes or catches. Their assessment of the situation is different. Their mobile lifestyles make it hard to attend drug distributions at static health centres, and they often do not hear about treatments that are meant to be available to them (e.g., free treatment for the water-borne disease, schistosomiasis) because little serious effort is made to include them. In such a context, it is easy to see why tensions came to a head so quickly during the first lockdown. One fisherman described the explicit targeting of fisherfolk in the following way: "Usually, a few minutes to the curfew time, armed men would be seen heading to landing sites to enforce the ban on movement of people … They would take cover and intercept whoever came late. Upon arrest, victims would be forced to remove their shirts, caned and asked to pay fines or taken to police cells … As time went on, this became a real business for the men in uniform." The strict enforcement of Presidential directives happened at a time when there were no reported cases of COVID-19 in Panyimur and COVID-19 was not perceived to be a serious affliction. The following comments capture this point: "it's a common 'flu that can't kill"; "it can be easily cured by chewing bitter fruits"; "COVID doesn't kill us Africans"; "it's a white man's disease; " "COVID is easy to treat"; "it is a foreign disease which affects those in the big towns"; "malaria is more serious than corona. It kills people"; "cholera killed over 50 people in a day [in 2009] … corona is less dangerous." Against this backdrop, there was widespread questioning of official directives, with many people expressing concern about militarised endeavours to secure health for political purposes. The following quotes, from a businessman, fisherman and teacher respectively, illustrate this point: "Museveni is using corona for doing his own things"; "the government is using corona to enrich itself …"; and "corona is considered a political disease because of the way security personnel were selective in enforcing guidelines." The use of the word security in this context refers not only to the UPDF, but also to the police, immigration officials and LDU'sall of whom were observed, to be working closely together. A local councillor, for example, described how money elicited from those paying (illegally) to cross the international border was placed in a collective pot, and shared out among security personnel at the end of each day.
From June 2020, restrictions gradually eased. The market re-opened, transport between districts resumed, and regulations on the number of people allowed to attend burials and marriages lifted. The LDUs were recalled along with some of the police and army personnel that had been sent to Panyimur from the district headquarters, and the situation was much calmer. Considerable efforts were made by the COVID-19 district task force to establish processes enabling health centres to prepare for future outbreaks of the virus. Although patients were required to bring their own gloves and face masks to the health centre, PPE continued to be unavailable and there was no attempt to enforce mask wearing or hand washing in other public places. In addition, schools and bars remained shut, a curfew was enforced, and movement across the DRC border was tightly monitored. The police and army continued to work closely together.
The second lockdown (from June-July 2021) was characterised by violence. Live rounds of ammunition were again fired into the air to clear the marketplace, and there were numerous reports of traders being beaten, arrested, and subsequently fined by local and district police forces. Public transport linking cities with Panyimur was effectively stopped, with the UPDF patrolling all roads leading to the town and near-by villages. Working closely with the police, they continued to enforce the curfew, and they also prohibited movement between districts and across borders, although exceptions were made for cargo trucks.
Although there were few confirmed cases of COVID-19 in the vicinity, there was increasing recognition of the potential seriousness of the affliction. A growing number of Alur words were used to refer to it including: mangota (a big, strong disease), kizi ngom (disease that affects the whole society), tho ngom (disease of the world), rutata (a very big disease) and rugwata (a tough disease). Others used the word corona to capture similar thoughts. The following quotes from a fisherman and a politician illustrate this point: "people have great fear of corona because they think it is a disease which has come to finish them off" and "corona is a dangerous disease … it has attacked all countries in the world". Perceptions of COVID -19 were undoubtedly shaped by information available on the news, WhatsApp and Facebook, including distressing accounts of people struggling to access health care in Kampala.
Responses to the new suite of measures imposed during lockdown varied but, in common with the first lockdown, this involved bypassing or subverting the regulations. However, this was not because there was no recognition of the potential seriousness of COVID-19, but rather because livelihoods had to be sustained to survive. Strategies varied and included: transporting goods at night-time on the back of motorcycles to 'beat the security roadblocks'. For daytime travel, motorcycle drivers used mobile phones to alert each other to the whereabouts of Ugandan soldiers. There are hundreds of motorcycle riders in Panyimur and they worked collectively by mapping out roadblocks, and identifying the best back roads. The strategy worked well, although they were occasionally caught by "men in uniform" and required to pay 50,000-100,000 shillings to prevent a motorbike being impounded. For those regularly crossing district borders in vehicles, payments were agreed in advance (usually 30,000-40,000 shillings), with soldiers often receiving payment directly to their phones in advance of their arrival at a roadblock. With some traders routinely passing through three or four roadblocks a day, the daily cost of travel could be considerable.
Crossing the international border presented a particular challenge. The border runs through two closely connected villages. Prior to the pandemic, no man's land was used as a football pitch and a place to dry fish before market day. The immigration post was run in a relaxed way: typically, one or two immigration officials sat in a mud-built hut and occasionally requested payment for the wooden pole to be lifted up so that a motorcycle could pass. Fisherfolk and traders crossed the border several times a day. Such movement continued during lockdown, albeit with modifications. This essentially entailed shifting movement to the Lake and using boats to dodge security checks by the UPDF and immigation officials stationed at the border post. It was less straightforward for those crossing the border by foot or motorcycle. Due to extensive deforestation in the area, visibility is good and it was relatively straightforward to constrain movement at informal crossings.
Panyimur trading centre has more than 10 nightclubs, 25 video halls and countless bars and guesthouses. The formal closure of these venues left people with little choice other than to subvert the regulations. For those with the space to do so, this involved closing the front of the building, and encouraging people to enter the premises around the back of the building and/or drinking discreetly outside. In common with the motorcycle riders, financial arrangements were made in advance with the police, LDU's and UPDF doing night patrols. Interestingly, authorities such as locally elected council chairmen participated in the process. The owners of a bar and guest house summed up the situation when they said: "the security people became important business partners to the extent of offering them drinks"; and "It has been a tough time … those who could not afford to pay [bribes] were put out of business." All of this was happening at a time when the UPDF's Fisheries Protection Unit (FPU) became actively involved in imposing nationwide regulations prohibiting fishing with boats less than 8 metres long. Initially introduced in 2017, and designed to prevent overfishing in the country's lakes, the FPU used the sustained military presence to impose the regulations. They seized and burnt fishing tackle and boats, with a local politician alleging that more than 300 boats had been burnt by March 2022. The overt destruction of local livelihoods contributed to a heightened sense of the armed forces creating fear and insecurity. One fisherman suggested that it was worse than a war.

Case study from Kasese district
Kasese district shares an international border with DRC. The Rwenzori mountain range forms part of the border. As with so many borders established under colonial rule in sub-Saharan Africa, the border separated groups of people who were socially and economically interconnected, and it made little sense locally. Related to this, there has been a long-standing movement to establish an independent Rwenzururu kingdom, involving the Bakonzo of Uganda and Banande of Congo (Titeca and Vlassenroot, 2012). The movement has been unsuccessful, and there have been several occasions in the 1990s and 2000s when the quest has led to open conflict and fatalities. The Allied Democratic Forces (ADF) which originated from Uganda and initially supported the idea of a kingdom, is currently based in DRC and remains a disruptive, occasionally violent and influential rebel group -openly hostile to Museveni's government.
Omusya (a pseudonym) village lies on the Ugandan side of the Uganda/DRC border. The Lhubiriha River runs along the western part of the village, acting as the international border. Most people are Bakonzo and speak Lukonzo. Subsistence agriculture characterises daily life, with beans, maize and cassava widely grown. Coffee beans and vanilla pods are also cultivated, and typically sold in nearby Ugandan and Congolese markets.
The military have had a long-standing presence in the area, partly because an internationally funded checkpoint was established at the edge of the village to monitor cross-border movements during the 2017-2020 Ebola outbreaks in Ituri and Kivu Provinces, DRC. Under the watchful eye of UPDF soldiers, the names of people crossing the border were logged in a ledger, with temperatures taken and handwashing required. In addition, UPDF soldiers regularly patrolled the border, including informal river crossings.
When the first lockdown was announced in March 2020, soldiers were already a familiar and feared presence. Typically, they had no connection with the district, spoke Swahili, English or Luganda, rather than the local language, and the majority of people avoided contact with them. In contrast to Panyimur, the UPDF did not shoot live rounds of ammunition into the air to assert their authority, (although it was reported that a priest was shot dead while travelling on a motorcycle in a near-by sub-county). They did, however, beat people for flouting the regulations. Typically, this was because they had been caught crossing the international border; moving around the village after curfew; or carrying passengers on the back of a motorcycle. A young man, reflecting on the violence meted out by UPDF soldiers said: "These soldiers have no mercy … they beat you like you are a snake … no-one can run there to stop them beating people unless you resemble them or you are their boss …" Although the UPDF were initially successful in restricting movement and activities, it was not long before concerns were articulated about the militarised response by formal and informal public authorities. There was widespread consensus that livelihoods could not be maintained if people did not cross the river to cultivate crops on land they owned or rented in DRC. Living in a context where there were no known cases of COVID-19 in the village or other parts of the sub-county, and COVID-19 was perceived as a 'disease of the radio', it made no sense for the army to jeopardise livelihoods in this way. Indeed, it was widely recognised that President Museveni was using COVID-19 to assert his political authority and undermine the influence of those seeking to replace him at the next election. As one participant said: "the big people are just using corona to keep themselves in government; " and another said: "[I] am realising that for him [Museveni], life is okay because he has all the food and money around himself. But for us, all the ways we can utilise to make money, and getting food are closed." Securing livelihoods in these borderland areas requires both pragmatism and innovation. Various strategies emerged over time, some of which involved resisting or challenging the presence of UPDF soldiers in subtle and subversive ways. In one instance, villagers were able to evoke the power of spirits residing in the river valley in such a way as to terrify soldiers and ensure that they did not monitor river crossings at night. One participant said: "the soldiers watching the river valley at night [do] not go beyond 9 o'clock. When it reaches this time, they fear the voice of the spirit of the river." Another said: "[t]he soldiers say that the spirit of the river instructs them to go away from the river because its valley cannot be watched using guns." Crossing the river during the night enabled people to move back and forth between Uganda and DRC. In so doing, they were able to cultivate their fields and bring produce back into Uganda. For those trading cattle, night-time crossings were not a viable option. Instead, they facilitated trade by involving Congolese and Ugandan soldiers. Typically, mobile money was sent to their phones on the understanding that they would oversee the movement of cattle across the border. Traders moving coffee, and other bulky items across the border faced similar issues. Reflecting on the restrictions in October 2020, one young man stated: "… the UPDF is busy wanting to arrest smugglers like me but we understand each other and we are working collaboratively". He went on to describe how he bought coffee from the DRC for 3500 shillings a kilo, and then sold it in Uganda for 5500 shillings a kilo, with the former sum including a payment to Congolese soldiers and the latter sum a payment to Ugandan soldiers.
In common with Panyimur, the first lockdown came to an end in June 2020 and some aspects of life resumed: the markets re-opened, and it was possible to move between districts. Over the coming months, restrictions on the numbers of people allowed to attend weddings and funerals lifted, with one villager commenting in December 2020 that "a wedding these days is like a market". However, the UPDF continued to monitor the international border and a night-time curfew was enforced. Although fewer soldiers were deployed (with numbers sometimes falling to two or three) and patrols along the river became less frequent, crossborder movements continued to be challenging. This was because villagers had to handle an unpredictable and volatile political situation on the Congolese side of the border. Numerous militia groups operate in the vicinity, including the ADF, who are known to be active and brutal. In August 2020, for example, they killed and beheaded a UPDF soldier in uniform, and in December 2021 there were terrifying accounts of villagers' relatives allegedly being beheaded. To complicate matters, the mai mai, nglima and M23 militia groups are also established fighting forces in the region, and to different degrees forge alliances with the Congolese army in opposition to the ADF.
Fearing the consequences of being ambushed by militia groups in search of food and arms, new strategies were devised to cross from Uganda to DRC in small groups at night-time. In one case, for example, a man rang clan members in the village, to provide updates of the whereabouts of local militia groups. When the path was clear, they crossed the river into DRCsomehow or other managing to return to Uganda with enough crops before dawn. Such acts of public mutuality reveal the way in which people help each other navigate multiple, unpredictable sources of instability, both longstanding and new. They also illustrate the counter-productive consequences of relying on the military to impose health containment strategies. Clearly, movement was not contained. On the contrary, it created new forms of mobility. To a degree, this was acknowledged by the commander who had imposed all regulations, when he invited villagers to a party to celebrate his promotion.
The militarised response to COVID-19 was widely perceived to be undermining endeavours to develop community-based health care. Village health team (VHT) members, for example, expressed disquiet throughout 2020 and 2021 that they had not received any official training about COVID-19 from health authorities within the sub-county, even though announcements were regularly being made over the radio that resources were being provided to support health care centres and related personnel. To quote a VHT: "Corona will not be managed well in this village because the people who are managing it are not the ones supposed to manage it. A soldier!" Another participant, reflecting on the fact that all the VHT members had been by-passeddespite successfully participating in other public health programmescommented: "Our respect is not considered." A local council chairman, expressing frustration with the enduring military presence, stated: "If epidemics are health issues, it is now time to follow normal health-approaches in disease containment including treating people, distributing vaccines and equipping health centres." The situation deteriorated with the second lockdown. While there were a few confirmed cases of COVID-19 in the village and other parts of the sub-county, and it was commonly stated that "corona is here" and could be "dangerous", there was far more concern about the renewed military presence. The number of soldiers deployed to the village increased, and cross-border movements were monitored in the way they had been during the first lockdown. For those wishing to make social visits to relatives, or to cultivate their fields, informal crossing points were sometimes used during the day. This typically involved giving money to Ugandan soldiers, but money was rarely given directly. Instead, it was mediated through a middleman and it was agreed, in advance, that bank notes would be dropped into the soldier's wellington boot when no-one was looking. Such practices ensured that if a soldier's supervisor turned up unexpectedly, it would not be found easily. For those trading cattle, similar strategies were used as before, with small sums of money being transferred electronically to the mobile phones of soldiers in advance of a crossing.
Unfortunately, the enduring everyday impacts of militarizing the response has ended up leaving people with a heightened sense of insecurity. This is partly because the economic impacts of lockdown have ended up being more disruptive than infection with COVID-19. The following quote, from a farmer living on the Uganda/DRC border elucidates the way in which security is linked to maintaining livelihoods.
"… in the absence of money, there is no way I can feel secure … being free from diseases means being safe … our health problem these days is malaria and typhoid. In the last few months, seven out of eight people in this home got malaria and only one did not … the medical prescriptions given in clinics require us to have money in our pockets to be able to buy the drugs prescribed. This is the safety I am talking about." He went on to say that it is hard to argue with official forms of authority such as the military, because "these military are like untouchable human beings … our military is authority itself."

Reflections from Kampala
Observations about the practice of militarised control measures for COVID-19 in the above two border locations were discussed by researchers with government officials. In large part, they were not surprised. The most revealing response came from a senior military officer on the National Task Force. He was aware that the restrictions enforced during, and following, the two lockdowns had created new forms of movement, both within districts such as Pakwach and Kasese and across international borders. While acknowledging that such movement could potentially have spread COVID-19, he did not think it had been influential. In contrast to the 2013-2016 Ebola outbreak in West Africa where the virus, he said, had moved from rural areas to urban conurbations (and back again), the primary goal in Uganda had been to prevent COVID-19 moving from densely populated cities to rural areas. Comparing Europe and the USA with Uganda, he also pointed out that the strategy devised by the National Task Force was underpinned by widespread awareness that the health system did not have sufficient staff, medicines or technologies to deal with a major outbreak. Given this, it was imperative that movement was halted from urban to rural areas. In his view, the strategy worked. Although he did not use the term collateral damage, the implication was clear.

Discussion
This article explores the ramifications of a militarised response to COVID-19 in two contrasting settings: a small town in Pakwach district, north-western Uganda, and a village in Kasese district, western Uganda. Both field sites are located on the Ugandan side of the Uganda/DRC border. Analysing events with a public authority lens, over a period of 18 months, provides a useful counterpoint to official narratives and usefully reveals socio-political dynamics that might otherwise remain hidden from view. In so doing, our research has shown the following.
First, the practice of health security varies within Uganda. In Panyimur, Pakwach district, soldiers are permanently located in Pakwach Town, which is more than 45 minutes drive away. Nevertheless, they arrived in the town within hours of the first lockdown being announced; and they asserted their authority by working closely with the permanently stationed police and immigration officials as well as the LDUs. In Omusya, Kasese district, there has been a permanent military presence since 2017, initially to prevent the transmission of Ebola from neighbouring DRC and subsequently to prevent the transmission of COVID-19. Soldiers regularly patrolled the international border, focusing on official Points of Entry as well as various informal crossing points along the river. The police and LDUs were minimally involved.
Second, the violence meted out by UPDF soldiers during lockdowns and curfews contributed to widespread questioning of official narratives at both fieldsites. The situation was not helped by the fact that the militarised response to COVID-19 happened when there were no official cases of COVID-19 in either place during the first lockdown, and very few confirmed cases thereafter. Although the potential seriousness of COVID-19 was recognised by the time the second lockdown was announced, no-oneto our knowledge -died from COVID-19 at these two fieldsites. Focusing on COVID-19 thus made little sense locally, especially when other infectious diseases remained endemic and caused significant morbidity, if not mortality.
Third, the militarised responses to COVID-19 have had more impact than the disease itself. The type of impact differs between sites. In Kasese, villagers navigate a complex space involving negotiations with the UPDF, the Congolese army and several militia groups. In Panyimur, the UPDF are working more closely with the police, LDUs and immigration officials, but militia groups in DRC or Uganda are not an issue. Nevertheless, at both field sites, the military scaled back movement from urban to rural areas; while simultaneously creating new rent-seeking opportunities for official public authorities, and new forms of crossborder movement.
This leads to a fourth and related point: militarised responses ended up accentuating, rather than diminishing, insecurity. In both places, people stated that they felt less safe because of enforcement measures. Livelihoods were rendered more precarious, and people came together to resist or bypass state regulations in mutually supportive ways. As time passed, soldiers stationed on the ground tended to shift away from exerting authority in ways that could be subverted and focused on what they could do more effectively, like extracting rents at border crossings. In more recent fieldwork, during April 2022, it was noted that such activities have become linked to enforcing vaccinations, often in ways that demonstrated their power rather than to ensure coverage of the population. Vaccines were used on traders at regulated crossing points each time they crossed, if proof of vaccination could not be produced. One man in Panyimur, who kept forgetting his vaccine card, was reported to have had the Johnson and Johnson vaccine twelve times.
What are the broader implications of these findings for furthering understandings of health security in Uganda and the engagement of armed forces in public health security more broadly? Prior to the outbreak of COVID-19, the Uganda Ministry of Health (2020) published a National Action Plan for Health Security, 2019-2024. The 101-page document outlined a comprehensive, "whole-of-government" (2019:12) approach to prevent, detect, and respond to public health emergencies. In many respects, the document elaborates global framings of public health security and clearly supports international endeavours to create a safer world (e.g., WHO, 2007;World Health Organization, 2022). The Ministry of Security is included as one of several partners contributing to disease surveillance activities, but explicit reference to the military and the security sector is omitted. Nonetheless, military deployment and policing are not explicitly excluded, and by promoting security as well-being, the document has effectively facilitated expansion of space for the security sector in the context of COVID-19. As observed elsewhere, epidemics have been used by states to assert centralised authority (e.g., Chigudu, 2016;Loewenson et al., 2020). Uganda is no exception. By embedding military activity in health security responses, President Museveni has strengthened unaccountable authoritarian rule.
There are troubling dimensions to this phenomenon. In the case of Uganda, it has been pointed out that the speed with which President Museveni responded to the pandemic in 2020, combined with the relatively low number of cases and fatalities, was referred to as one of several African 'success' stories (MacGregor et al. In Press). Indeed, data analysed by the Lancet Commission in August 2020 suggested that Uganda had been more effective at containing COVID-19 than any other African country, and it was ranked 10th out of 191 nations worldwide (Lancet Commission, cited by Cheeseman, 2021) ). Recognising this 'achievement', the British Medical Journal and the National Health Service Health Education Programme awarded President Museveni a medal for his "extraordinary leadership in pandemic management" (Kasibwe, 2022).
Assessing success -or failure -in contexts where diagnostic facilities are limited and surveillance systems patchy is challenging. Nevertheless, juxtaposing the findings presented in this article with epidemiological indicators of Uganda's 'success', suggests it is misleading to rely on so few indicators. To do so, overlooks a crucial point: it is possible that the relatively small number of confirmed cases of COVID-19 and fatalities had relatively little to do with the restrictions imposed during and following the two lockdowns, and much more to do with factors such as: Uganda's youthful population (Uganda Bureau of Statistics (2020) estimated that 50% of the population are less than 17 years old and 1.6% are 70 or more years old); the smaller proportion of people with conditions such as diabetes and hypertension, associated with severe COVID-19; and the fact that work and social activities often take place outside, especially among rural populations. While there can be little doubt that movement was curtailed (particularly between urban and rural areas), the data presented in this article suggest that it was not prevented altogether. Instead, lockdown created new forms of movement. Diverse strategies were devised 'to beat the men in uniform' and continue economic activitiesoften across international borders.
A second, and rather different point, is that endeavours to assess the merits or otherwise, of using the armed forces to contain transmission of COVID-19 typically ignore the broader socio-political ramifications of such an authoritarian approach. In Uganda, the militarised response enabled official public authorities to use the concerns of the state to enhance their authoritymost notably the UPDF and police, but also the Fisheries Protection Unit tasked with surveilling Lake Albert. This, in turn, led to a situation whereby inhabitants in both Omusya and Panyimur created new modes of mutuality to resist or subvert the regulations being enforced. Indeed, if they had not found ways to assist each other, they would have struggled to fish, cultivate their crops, feed their children and/or generate income for health care. Such modes of mutuality resonate with those reported by Parker et al. (2019aParker et al. ( , 2019b during the 2014-2016 Ebola outbreak in Sierra Leone and  during a COVID-19 outbreak in Bukavu and Goma, DRC. This is not to suggest that soldiers cannot be helpful in responding to epidemics. For example, it has been reported that Ugandan military medics were influential in improving the treatment of Ebola patients in Sierra Leone (personal communications in 2017 and 2022). Nevertheless, the above-mentioned decision by the Lancet to rank Uganda the 10th best-performing country worldwide, and the award of a prize to President Museveni for his exemplary leadership during the pandemic highlights a concerning trend: violent acts are occurring in the name of 'health security'. In the past, such acts would have been castigated as human rights abuses. The fact that they are often condoned and valorised reflects a shift towards more authoritarian forms of governance, even in some countries in the Global North, where democracy and human rights have previously been idealised. This resonates with Fisher and Anderson's (2015) analysis, which foregrounds the way in which African Presidents from Chad, Ethiopia, Rwanda and Uganda use the language of securitisation to foreclose debate and enhance their authority in realms beyond public health.

Conclusion
There is a long history of armed forces being mobilised to contain outbreaks of infectious disease in African contexts. However, when the UN Security Council passed Resolution 2177 much changed. It not only legitimised the use of force during the West African Ebola outbreak but also increased the political space for armed forces to be used to control other African epidemics. In Uganda, the new turn in global framings of public health security was the context in which national policies occurred. Whatever their intentions, these policies have enhanced President Museveni's capacity to actively hollow out democratic processes and strengthen authoritarian rule. Drawing on ethnographic fieldwork carried out in two borderland areas, this article reveals the way in which current global and national discourses have shaped the day-to-day practice of health security. In both places, involving the military created a heightened sense of insecurity. Risks had to be taken to survive, and livelihoods were rendered more precarious. Public mutuality often resisted or subverted official directives, as people came together to assist each other. Overall, it is argued that internationally encouraged COVID-19 control measures have meant that public health, the armed forces, political security, and the institutionalisation of social control are now inextricably interlinked. in Uganda, such trends have opposite effects to those imagined when the word security was used in the Preamble of the World Health Organization. Meanwhile, the degree to which the spread of COVID-19 has actually been contained in Uganda by military involvement remains moot.