Cabo Ligado Monthly: July 2021
July At A Glance
Vital Stats
ACLED records 38 organized political violence events in July, resulting in 134 fatalities
Events took place in Ibo, Macomia, Mocimboa da Praia, Montepuez, Muidumbe, Nangade, and Palma districts
Vital Trends
Rwandan troops arrived in Cabo Delgado in July and quickly entered the fray; they were involved in 10 recorded organized political violence events resulting in 73 reported fatalities
Violence in Mocimboa da Praia district jumped sharply as Rwandan and Mozambican troops undertook an offensive that ultimately resulted in the re-taking of Mocimboa da Praia town in early August
Fighting also continued in Palma district, as Rwandan and Mozambican troops worked to clear the areas around Palma town and the route from Palma to Nangade of insurgents
In This Report
Analysis of KiSwahili-language Islamic State propaganda streams and their relationship to the Cabo Delgado conflict
Explanation of the origins and functions of pro-government local militias in Cabo Delgado
Examination of the combat record and political position of Rwanda in the Cabo Delgado conflict
Update on international involvement in Cabo Delgado with a focus on how costs and responsibilities will be divided up between SADC member states in the SADC Standby Force mission
July Situation Summary
In July 2021, the Cabo Delgado conflict became internationalized to an unprecedented extent. Rwandan troops entered the fray at the Mozambican government’s behest, sending police and military forces to assist the Mozambican counterinsurgency effort. The Southern African Development Community (SADC) began to deploy troops for its Standby Force mission to Mozambique, which adopted the acronym SAMIM. On the insurgent side, the Islamic State (IS) dramatically increased the pace of claims it issued of attacks in Cabo Delgado, seemingly in response to the international interventions.
The most immediate consequence of the increased internationalization of the conflict was increased violence between armed groups. Mozambican forces, Rwandan forces, or joint contingents from both countries fought with insurgents in 22 recorded armed clashes, resulting in 95 reported fatalities. Rwandan troops, whose combat record will be discussed in depth in this report, were in action in Palma, Muidumbe, and Mocimboa da Praia districts. Much of the fighting they were involved in came during a multi-week offensive aimed at recovering Mocimboa da Praia town, which succeeded in early August. SAMIM troops did not see combat in July, and the exact role they will play in the overall counterinsurgency strategy was not clear by the end of the month.
Another consequence of internationalization was an increase in international actors making public pronouncements about the conflict. The Rwandan defense ministry began offering detailed press briefings about Rwandan involvement in Cabo Delgado, temporarily displacing the Mozambican government as the main source of information about events on the front line. Those briefings offered detailed descriptions of engagements between Rwandan troops and insurgents, as well as inventories of materiel that Rwandan troops captured from insurgents. At the same time, however, it allowed the Rwandan government to portray its intervention as essentially costless, claiming that Rwanda Defence Force (RDF) troops have suffered no fatalities. That claim—unlikely on its face given the amount of combat the RDF saw in July—contradicts multiple reports.
IS began releasing attack claims for Cabo Delgado again in July, after about three months without a claim. The renewed media activity seemed timed to correspond with the arrival of Rwandan troops in Mozambique and SADC’s decision to send a force of its own. Yet the claims themselves made no mention of foreign forces and largely shied away from describing action on the main axis of Rwanda’s advance toward Mocimboa da Praia. Instead, they focused on smaller engagements elsewhere in Cabo Delgado where IS could claim that insurgents were still making progress despite widely reported counterinsurgency successes. IS media only referenced foreign intervention in Mozambique at the beginning of August, and then only glancingly.
Online Islamic State Propaganda
After months of little coverage, and even less audio-visual material from the field, Cabo Delgado is now receiving some attention in online information networks – both in IS central media apparatus, and in an array of aligned social media accounts. The attack on Palma in March led to a spike in online activity related to the conflict, while the arrival of first the RDF and then SAMIM has seen sustained coverage of events across channels controlled by the Islamic State, and others sympathetic to it, over the past month. Information on the conflict in Cabo Delgado through these channels lacks detail, and is backed by no significant audio-visual material. Most of the material from IS-aligned accounts relies on the same open source data on which mainstream analysts rely. Nevertheless, it creates the basis for significant flows of propaganda in the Kiswahili language, targeting East Africa.
‘Official’ IS reporting of incidents follows a set pattern. Brief incident reports from across the provinces, presented as text in images, are distributed through channels on social media platforms such as Telegram and Element, in Arabic. These are then collated and posted on sites such as Archive or Justpaste.it, posts that are again disseminated on Telegram. Finally, incident reports are aggregated into reports from the provinces in the IS weekly newsletter, Al Naba. They also provide the raw data for Al Naba’s weekly ‘Soldiers’ Harvest’ infographic. A recent development has been a Kiswahili language podcast, distributed weekly through Facebook and other channels, based on weekly incident reports but also including ideological segments. It is not clear if this is supported directly by IS.
Over the past two years, there have been varying strategies used to produce IS-supporting propaganda targeting East Africa, across open platforms such as Facebook, Instagram, and YouTube, and closed messaging ones such as WhatsApp, Telegram, and Element. Channels change as platforms are seen to be more or less successful in identifying and removing material. For the past two years, Facebook has been used extensively to target East African audiences in regional languages. Telegram remains in use, though in recent months some channels targeting East and Central African audiences appear to have been removed.
The most successful home-grown media operations have come from Islamic State Central Africa Province in Democratic Republic of Congo (ISCAP – aka Allied Democratic Forces or Madina at Tauheed Wau Mujahedeen). These have provided timely and generally accurate reports of operations, accompanied with battlefield images and video for use by the central media apparatus of IS. Alongside this, they produce a steady output of propaganda in a range of languages, including Kiswahili, Kinyaruanda, Luganda, and French. Such material appears across all platforms. Footage from an attack distributed on IS channels on 6 August has since appeared on one at least one Facebook page, and is circulating widely in Tanzania on WhatsApp.
By contrast, the link between IS central media apparatus and Cabo Delgado is weak. Heretofore, the fitful appearance of incident reports suggested no sustained relationship. Similarly, there has been no sustained production of propaganda material by the insurgents themselves, apart from a flurry of clips in May 2020. Nevertheless, the limited number of reports on Telegram -- Cabo Ligado has tallied six since 13 July 2021 -- has been enough for prominent coverage in each of the four issues of Al Naba published since 15 July 2021. The editorial in the 5 August edition predictably focused on intervention by the “African Alliance,” drawing comparison between RDF and SAMIM and the more limited engagement of Uganda and Kenya in Eastern DRC in May 2021.
Though incident reports and Al Naba are published in Arabic, IS supporting initiatives regularly adapt and translate content. In the first half of this year, incident reports, with Kiswahili text in images, were consistently produced and distributed via Telegram and Element, though this appears to have stopped or moved off channels being monitored. More prolific -- and reaching a much larger audience -- are Facebook accounts. Limits on who can view posts, as well as restrictive access to the underlying mechanisms of the platform, make it almost impossible to monitor propaganda trends on Facebook from outside the platform. However, the small number of mostly open accounts give a flavor of the type of messages being disseminated, and a sense of the broad target audience. They also indicate that this platform has considerable reach.
Accounts sympathetic to IS are of concern for four reasons. Firstly, the production and dissemination of such content is consistent. One Kiswahili language account has posted 16 reports related directly to Cabo Delgado over the past two months out of a total of 102 public posts. These comprise reports on the actions of IS and its affiliates, as well as lengthier works across multiple posts on subjects such as the history of IS. The account has been active since at least December 2020.
Secondly, the reach of these accounts is significant. Each of the noted account’s 16 posts on Cabo Delgado were shared between two and 13 times, attracted between 19 and 72 comments, and garnered between 43 and 94 reactions. One’s reaction or comment on a public post is visible to one’s friends, so the potential reach of the original posts is significant, though not exactly quantifiable given varying privacy restrictions people may put on personal accounts.
Thirdly, these accounts are not reliant on direct communication from the field. Reports and analysis from mainstream media and other outlets monitoring the conflict consistently provide the basis of posts. In this way, propaganda based on events in Cabo Delgado, and produced in Kiswahili and other languages in the region, presents a regional threat.
Finally, the consistent presence of such accounts, particularly on Facebook, illustrate weak controls by platforms themselves, and illustrate the continuing efforts required by initiatives such as the industry’s Global Internet Forum to Counter Terrorism.
The Role of Local Militias in the Conflict in Cabo Delgado
Throughout the conflict in Cabo Delgado, civilians have quite frequently been reported to be joining the armed groups that have been carrying out attacks across the province. While some join the ranks of the insurgents voluntarily, others are forced to join through coerced recruitment. However, little has been said about the civilians who decide to take up arms to defend their villages and towns from insurgents. These newly-formed armed groups are generally known as local militias or local forces.
According to sources interviewed by Cabo Ligado, the first appearances of local militias in the conflict date back to the first quarter of 2020, in Muidumbe district. Locals there feared being the insurgents’ next target, in the wake of the insurgency’s seizure of Quissanga town and successive attacks on Mocímboa da Praia town before its capture by the insurgents in August 2020. From then on, local forces have played a crucial role in the conflict.
The local militias are made up of veterans of Mozambique’s liberation war and soldiers demobilized after the country’s civil war, as well as those who have completed compulsory military service, people who have suffered at the hands of insurgents, and others. Although they only started to join the conflict in Cabo Delgado in 2020, the war veterans had indicated well before then their willingness to take up arms again and make use of their combat experience to stop attacks in Cabo Delgado. However, their appeal had not been promptly accepted by authorities.
Local militias were eventually created by the government to respond to a number of challenges faced by Mozambican defence and security forces. The first challenge was the lack of geographical knowledge of the terrain. Unarmed civilians were included in raids by Mozambican security forces to serve as guides, but after repeated attacks by insurgents, they saw the need to carry arms for self-defence. In Nangade district, the local forces became part of the joint forces comprising civilian protection police, border guards, and the military, which are responsible for patrolling in conflict zones. A second challenge was the lack of capacity of the Mozambican security forces to provide territorial coverage in several towns and villages at the provincial level. Another factor driving the formation of local militias was popular pressure on the government after successive successful pursuits and attacks on insurgents by a group of men led by a former combatant known as Old Naveta from Mueda. In one of the attacks led by Naveta, in October 2020, local militias were rumored to have killed around 270 insurgents.
The local militias have an independent command, and receive only weapons and uniforms from government authorities. Sources say that the leaders of the local militias have access to communication channels with the military command, but its members act autonomously. The fact that they act with a certain impunity and lack of accountability to the state can lead to human rights violations. As the state has no direct control over local forces, this can lead to the proliferation of weapons beyond the control of government authorities which could be a problem in the post-conflict era.
Members of the local militias do not receive a salary. War veterans receive their regular pensions, but these pensions are not increased by virtue of the veterans serving in militias. Sources living in Palma have said that due to the high cost of living, before the last attack on Palma, the local militia escorted food trucks from Mueda to Palma (via Pundanhar) and back in exchange for payment.
Although the local militias work in coordination with state security forces in patrolling the conflict zones, the relationship between the two forces has sometimes been marked by suspicion and mistrust. On the one hand, state forces accuse the local forces of being unprepared and lacking military training (since the volunteers do not receive adequate training), limiting themselves to learning how to use AKM-type weapons. The military complains that local militias, when faced with insurgents, decide to attack without first drafting a plan.
Meanwhile, the local militias claim that they do not know the real enemy, since the insurgents wear uniforms similar to those of the state security forces. The militias also complain that the military does not allow them to attack the enemy. Suspicions are also common among the local militias. In several incidents, local militias have suspected the military of killing and looting civilian property, and of collaborating with insurgents. One of the most recent incidents was the 12 June killing of seven civilians by beheading in Litamanda, Macomia district, which locals first suspected to have been done by insurgents. They later accused the military of being behind the killings.
During the recent attacks on insurgents by Mozambican and Rwandan forces, which resulted in the capture of several villages and towns, including Awasse and Mocímboa da Praia, local militias played a leading role, according to some sources, thanks to their extensive knowledge of the terrain. The local militias -- drawn from the ranks of local civilians and deeply embedded within local networks -- have also built up trust among local civilians. When joint Rwandan-Mozambican patrols include local militia members, therefore, the joint forces are better able to interact with local civilians.
Rwanda’s Combat Record
In July, Rwandan forces entered the Cabo Delgado conflict with a contingent of 1,000 personnel -- roughly 300 police officers and 700 soldiers. They quickly entered the fight, deploying in Palma, Nangade, and Mueda districts. They then moved from Mueda through Muidumbe and into Mocimboa da Praia district.
On the battlefield, Rwandan forces have made an immediate impact. After a year in which periodic attempts by Mozambican forces to push through the N380 corridor from Mueda to Mocimboa da Praia town resulted in repeated failure and occasional disaster, Rwandans were able to retake the strategic towns of Diaca, Awasse, and, eventually, Mocimboa da Praia itself in under a month. These remarkable early returns are notable both for what they show about the deficiencies among Mozambique’s security forces and for how we should expect the insurgency to respond to these new foes.
From a military standpoint, the most remarkable thing about the Rwandan troops’ success is how little their operational concept differs from that of their Mozambican counterparts. Their predecessors in intervention, the Wagner Group, tried to introduce a new operational concept rooted in the use of drones and indirect fires. That concept, drawn from Wagner’s experience in the wide open spaces of Ukraine and Libya, floundered in Cabo Delgado’s dense forests. The Rwanda Defence Force, by contrast, uses the same direct fires, armored columns, and dismounted infantry that Mozambican troops have relied on throughout the Cabo Delgado conflict. They bring no new aerial assets to the conflict and remain as reliant on major roads as Mozambican forces have been -- a weakness that has long hamstrung Mozambican operations. Despite this, Rwandans have had much greater battlefield success than their Mozambican counterparts.
Early reports suggest at least one key reason for this divergence: that Rwandan troops are more professional in the face of ambush than Mozambican troops. In battles for the N380 corridor, the most common insurgent approach has been to set ambushes for Mozambican troops as they travel along a predictable, heavily wooded route. Mozambican troops invariably retreated in the wake of these ambushes, sometimes leaving behind valuable equipment. As Rwandan troops advanced along the N380, they also faced insurgent ambushes, such as the incidents on 28 July near Awasse. Rather than fall back, however, Rwandan troops had the training, leadership, and experience to assault through the ambush and win these engagements. Winning ambush encounters neutralized the insurgents’ main strategy for defending the approach to Mocimboa da Praia.
The Rwandan intervention has also proved to be remarkably politically astute, to this point at least, in its presentation to local audiences in the conflict zone. Approval of the Rwandan deployment appears quite high. Civilians in Nangade and Palma districts, where Rwandan troops have been deployed and interact with civilians frequently, report highly favorable opinions of the intervening force. Civilians in Palma district who remain outside the civilian safe zone at Quitunda say that they feel confident in the Rwandans’ ability to differentiate them from insurgents—a remarkable feat, given the language barrier. Indeed, it appears that direct civilian outreach from Rwandan troops has been at least partially successful. One Palma civilian told Cabo Ligado that the Rwandans “are gaining our trust and we are loyal to them,” adding that the foreign soldiers are there to “bring peace.”
In Nangade, civilians reported feeling “reassured” by the Rwandan presence, but some did not have the same feelings about an impending deployment by SAMIM. Rumors that South Africans are involved in the insurgency are nearly as old as the insurgency itself, and some suggested that it would be difficult to trust South African troops as a result. In this respect, Rwanda’s physical and cultural distance from the conflict is an asset, rather than a liability.
Despite concerns among segments of both Frelimo and Mozambican civil society that Rwandan intervention would be seen in Mozambique as a threat to Mozambican sovereignty, the civilians closest to the Rwandan deployment seem to see it more through the lens of solidarity. Mozambican president Filipe Nyusi has invoked the history of solidarity among African independence movements as an explanation for the Rwandan intervention, and it is a history that is intimately familiar to many in Cabo Delgado. The province served as both a key base area for Mozambique’s independence war, undertaken with crucial support from neighboring Tanzania, and as a training site for Ugandan fighters in the fight against Idi Amin’s regime. Even if Rwandan troops are foreign to Cabo Delgado, the concept of intra-African state military solidarity is not.
SADC Developments
SAMIM has been officially announced, but details on its deployments and mandate remain sketchy. Although the mission is only authorized for three months, at this juncture, it is already clear that this will likely be extended, contingent on conditions on the ground and available funding. SADC has had a slow start out of the gate compared with the rapid deployment of Rwanda’s security forces, who were on the ground within days of announcing their deployment and involved in combat operations not long after that.
The RDF deployment had created tension in some quarters. South Africa’s (now former) defence minister, Nosiviwe Mapisa-Nqakula, complained to South Africa’s public broadcaster on 9 July that it was regrettable that the RDF deployed ahead of SADC, suggesting the move contradicted SADC’s collective agreement. Mapisa-Nkaqula was not, however, representing an official SADC posture as some headlines suggested. Mapisa-Nkaqula subsequently withdrew the comments following presidential censure and after Mozambican president Filipe Nyusi pushed back, arguing SADC respected Mozambican sovereignty, that SADC was not in charge of security operations, and that the regional body had in fact consented to Mozambique’s bilateral security arrangements with Rwanda. Later in July, Nyusi insisted foreign support would remain under Mozambique control and direction. Mapisa-Nqakula was subsequently sacked from her position by South African president Cyril Ramaphosa as part of a major cabinet reshuffle in early August.
On 8 July, SADC announced that SAMIM deployment would commence on 15 July for a three-month period until 15 October. This would first require a “status of forces” agreement, which provides official authorization for member state deployment. The agreement had been expected by 14 July, but by the night of the 13 July had still not been signed, prompting speculation that Maputo would use this to delay SADC’s deployment further, in line with its general resistance to SADC involvement. However, the following day, Nyusi reportedly approved the deployment, opening the door for SADC. On 16 July, only a day behind schedule, SADC Executive Secretary Stergomena Tax presented the “instruments of authority” to the SAMIM Force Commander, Major General Xolani Mankayi, an African National Congress war veteran who served as Commander of the South African National Defence Force (SANDF) Peace Mission to Burundi, and has been Commander of the SANDF 43 South African Brigade since 2015.
Mankayi will work with Mpho G. Molomo, Secretary for Political Affairs, Office of the Presidency, in Botswana in his capacity as the Special Representative of the Chairperson of the Organ on Politics, Defence and Security Cooperation. Molomo is effectively the civilian head of SAMIM, reporting to the Organ chair. His position may be short lived, however, if the incoming chair, Cyril Ramaphosa, decides he wants his own man in that position. Although Molomo was only appointed on 14 July, there is a case to be made for continuity by retaining him in situ.
The details of both the military and civilian aspects of SAMIM’s mandate are still unclear. The military section presents its role in terms so broad as to defy prediction of what actions the mission will undertake. According to correspondence from SADC to the United Nations Secretary General, SAMIM’s military contingent will provide “support (to) the Republic of Mozambique in the fight against acts of terrorism and extremist violence, and support the country in restoring the rule of law in the affected areas of Cabo Delgado province.”
But detail on the civilian mandate of SAMIM is even sketchier than its security objectives. While SADC Standby Forces were generically envisaged to include a policing element, as most UN missions are, SADC’s Technical Assessment Mission in Mozambique mentioned a policing component only in passing; detail of its role and competencies are unclear. There is also the wider question of how SAMIM will support Mozambique’s wider efforts to address human security challenges, beyond the immediate humanitarian and stabilisation priorities. Options for rebuilding institutional capacity and service delivery will require investment in participation and community dialogue.
SADC has provided limited detail on the actual deployment of troops. Details from media reports provide some insights, but also raise further questions on who is coming and at what stage. For financial and political reasons, it is highly unlikely that 3,000 troops will be deployed as had been recommended by SADC’s Technical Assessment Mission in April.
So far, it appears the Troop Contributing Countries (TCCs) are South Africa, Botswana, Tanzania, Zimbabwe, Angola, and Lesotho. There are questions as to whether Malawi or Namibia will contribute.
SANDF and the Botswana Defence Force deployed advanced teams and equipment to Pemba on 19 July. This included special forces components. According to South African security expert Darren Oliver, “their main mission is to gather intelligence, conduct reconnaissance, advise the Mozambican military, and prepare command and control structures for a potential deployment of a full SADC brigade.”
On 23 July, Ramaphosa authorized up to 1,495 SANDF personnel for deployment at an estimated cost of almost $68 million. Several media outlets interpreted the 1,495 as the number of troops that would be sent, but this is not the case as it is only the number authorized and no final deployment announcement has been made. By 31 July, SANDF’s 43 South African Brigade vehicles crossed into Mozambique at Ressano Garcia on the long drive to Pemba under escort of Mozambican military police. The Brigade components are bespoke to the task at hand; it has been honing its skills in jungle training earlier this year which will be important for the difficult conditions in Cabo Delgado. In addition, the SANDF has deployed the SAS Makhanda, a Warrior-class strike craft of the South African Navy, currently configured as an Offshore Patrol Vessel. It arrived in Pemba as part of the SADC intervention force in the first week of August
On 26 July, Botswana reportedly sent 296 troops. Their accompanying armored personnel carriers and other heavy equipment were seen on 30 July on the road at Gondola, Manica Province, heading north.
On 27 July, Angola’s parliament approved the deployment of 20 advisers and a transport aircraft to join SAMIM. A budget of $575,000 has been allocated to cover their three month placement.
Following the June summit, Zimbabwean independent media was speculating that Zimbabwe was poised to deploy combat troops. On 29 July, Defence Minister Oppah Muchinguri approved the deployment of 304 trainers. It is unlikely that this number will be deployed as a single group, but rather in some kind of rotation.
Further detail on deployments has followed in early August, but without official confirmation. This includes information about a small contingent from Lesotho and several hundred from Tanzania, contradicting earlier reports that they would not be contributing.
SAMIM faces an unprecedented and complex set of challenges, both in terms of the nature of the security challenge, conditions on the ground (i.e. limited knowledge of the terrain and major intelligence deficits) and longer term social, economic and political challenges. Financing this mission remains a major challenge. On 28 June, the SADC Council of Ministers approved a $12 million budget to be drawn from the contingency fund. A further $7 million contribution by member states was predicted to be made available on 9 July -- other media suggests the amount was $11.8 million. The initial allocations are clearly insufficient, even if South African underwrites the bulk of the bill. Any extension of the mission beyond October will require external financing.
SADC’s Standby Force has also limited preparedness for such a mission. SADC has limited counterinsurgency capacity. It also has not developed competencies in terms of the regional counterterrorism strategy that was adopted in 2015, despite the recognition at that point that insurgencies in other parts of the continent, such as Boko Haram in Nigeria and Al-Shabaab in Somalia “could spill over to other parts of Africa, posing a potentially serious security and humanitarian threat.” The current mission will test the extent to which relevant preparations to prioritize and address such threats are in place. The insurgency has certainly resuscitated SADC interest in implementing counterterrorism measures as part of efforts “towards forging a holistic approach for sustenance of security at both national and regional levels.” At the Council of Ministers meeting on 7 July, Executive Secretary Stergomena Tax highlighted the importance of operationalizing the Regional Counter Terrorism Centre as part of the SADC Regional Counter Terrorism Strategy.
However, the task at hand goes beyond the strictures of counterterrorism if a sustainable solution is to be found. It remains to be seen if and how SADC will be able to work with Mozambique to address the longer term and underlying drivers of the conflict. Both Maputo and the SADC Secretariat are ill-equipped in terms of institutional capacity to tackle these issues on a range of fronts from development to dialogue and it remains to be seen how these considerations will be factored into their longer term strategic approach.