September at a Glance

Vital Stats

• ACLED recorded 47 organized political violence events in September 2020, resulting in 92 fatalities

• The deadliest set of incidents were insurgent attacks in Palma district, where 37 people were killed as insurgents attempted to cut off the road between Palma town and Mueda. Another 29 were killed in Macomia district, most during an apparent insurgent offensive there at the end of the month

• Events took place in Macomia, Mocimboa da Praia, Muidumbe, Nangade, Palma, and Quissanga districts. Palma and Macomia district were where the bulk of incidents took place

Vital Trends

• There were increased incidents on islands off the coast of Cabo Delgado in September, reflecting the insurgency’s increased access to water transport following the occupation of Mocimboa da Praia

• Events were more geographically diffuse in September, with insurgents re-extending their reach southward into Quissanga district and running simultaneous offensives in Palma and Macomia districts

• The security progress the government had made in Quissanga and western Macomia district dissipated in September, as both areas were hit with the first insurgent attacks in months. The areas are still more secure than eastern Macomia district or points north

In This Report

• Analysis of the coming food crisis in Cabo Delgado as it relates to the insurgency

• A review of the evidence for and against ethnicity as a main driver of the insurgency

• An update on Mozambique’s negotiations with the EU and SADC about future intervention in Cabo Delgado

September Situation Summary

Insurgents spent September 2020 expanding their areas of control in the northern districts of the conflict zone, while gathering resources and sowing terror in southern districts that the government once hoped were safe from future attacks. 

After months away, insurgents returned to attack Quissanga district on 22 September. Occupying BIlibiza, a town insurgents had struck multiple times during their Quissanga offensive in early 2020, the attackers accused civilians of disloyalty to the insurgency during the period of government control. They then stole phones and killed two civilians. Two subsequent attacks in Quissanga underlined how weak the government’s security guarantee there was. 

Just to the north, a more concerted insurgent offensive engulfed the Mucojo administrative post of eastern Macomia district. Insurgents launched attacks in Mucojo throughout the month, starting on 1 September with an insurgent raid on Nambo in which two civilians were killed and houses burned. The attacks increased in frequency and intensity toward the end of the month, however, with the occupation of Pangane and Mucojo town. Government troops were killed in both of those engagements.

Insurgents in the northern part of the conflict zone maintained their presence around Mocimboa da Praia town and attempted to expand the areas of the northeast cut off from government control. In Mocimboa da Praia district, insurgents sabotaged electrical infrastructure and then attacked national electricity company employees who came to repair the damage. In Palma district, insurgents began an extended campaign to prevent traffic on the road between Palma and Mueda, which is the last overland route between Palma and the rest of Mozambique that remains open following the insurgent occupation of Mocimboa da Praia. That campaign killed at least 26 civilians and led to the road being closed for a significant period.

Further evidence of human rights abuses by Mozambican troops also came to light in September. Amnesty International confirmed the authenticity of five videos showing Mozambican security forces abusing and killing prisoners, and a video appeared on social media of soldiers beating and then executing a naked woman on a Mocimboa da Praia district road. The Mozambican government has denied that any of the videos are real. 

Despite the embarrassment caused by the videos, the Mozambican government pressed on with their attempts to garner international support for their counterinsurgency effort. In a letter to the European Union, Mozambique requested assistance in the form of training for Mozambican troops, medical equipment, logistical support, and humanitarian aid. The EU confirmed that some support would be forthcoming, but, as this month’s report makes clear, the details are still very much up in the air.

The Insurgency and the Coming Food Crisis

The Famine Early Warning System reports that Cabo Delgado is experiencing an “atypically early lean season” this year as a result of the conflict. This period, in which food stores dwindle as new crops are planted, would normally run from roughly December to March. Instead, in some parts of the province, the lean season is already beginning. Staple food prices were up 20-43% over five-year averages across the country in August. They are expected to rise further in Cabo Delgado as the conflict disrupts subsistence agriculture and more people become dependent on market-bought food. In short, coastal Cabo Delgado is about to enter an extended food crisis.

Civilians will be most affected by the food shortage, but insurgents will not escape the crisis. In contrast to more traditional rural insurgencies, which try to separate agricultural producers from the food market and use their production to feed insurgents and supportive civilians, the Cabo Delgado insurgents have driven farmers away from their base areas. As a result, insurgents rely on local supply streams to maintain their bases, and are thus subject to local fluctuations in the food economy. As prices have risen in anticipation of the lean season, there have already been two effects: increased food raids by insurgents to acquire food outside the bounds of the market; and reports of unknown buyers — potentially connected to the insurgency — offering high prices for food and other supplies.

In September, ACLED recorded eight confirmed incidents of food gathering during insurgent attacks, nearly all in Macomia district. In August, only one such incident was recorded. These numbers come with two significant caveats. The first is that they are both almost certainly undercounts, given that stolen food is sometimes considered a minor consequence of attacks in which people are killed or buildings are destroyed. It thus often goes unreported. The second is that the ongoing occupation of Mocimboa da Praia town, and the lack of information coming out of the occupied zone, make it hard to measure how much food gathering is going on in the northern part of province around Mocimboa da Praia. However, it is clear that insurgent food gathering operations are increasing, at least in Macomia district. 

As the lean season becomes leaner, the nature of insurgent food gathering operations is likely to change due to the arrival of the rainy season, which begins in November. The rainy season restrains freedom of movement, making it very difficult for insurgents to execute a supply strategy based on frequent small raids to acquire food. The worsened travel conditions offer insurgents at least three options for pursuing resupply, all of which they are likely to pursue. The first is to trade smaller, more frequent food raids for larger, harsher, and less frequent raids. The number of insurgent attacks on civilians may decline as the difficulty of transport increases, but the attacks that do happen are likely to be quite severe. Insurgents will need to maximize the results of each operation both in terms of supplies gathered and in terms of reminding civilians of their coercive power.

Second, insurgents may utilize the coastal areas they have recently focused on occupying to acquire local fish catches and supplies from outside Cabo Delgado. Between the ongoing occupation of Mocimboa da Praia town, which gave insurgents access to a small fleet of motorboats, and the insurgency’s recent focus on attacks in coastal Macomia district, the insurgency could enter the rainy season with a good deal of coastline under its control. With sea-based trading well established in Cabo Delgado, both with foreign traders and, to a lesser extent, with domestic sources to the south, insurgents could turn to the coastline as a key element in their supply line in a way that they have not done thus far.

Finally, insurgents may expand their operations on the grey market, spending cash reserves to access the market for food in Cabo Delgado. This approach is limited by the size of the insurgency’s cash reserves. However, the repeated occupations of towns with banks over the course of the year may mean that insurgents still have money to spend. These transactions, which would likely involve a significant premium for the seller given the danger of being caught selling to insurgents, could place further strain on the civilian market by driving the market rate for food supplies higher.

In the end, the insurgency’s forced displacement strategy and the level of isolation it has created may end up backfiring during this lean season. Of the eight food raids recorded in September, none came after battles and only one involved any civilians coming to physical harm. Food gathering, in other words, was quite simple for raiding parties. However, as resources become more dear, more displaced people gather near government controlled areas, and the government expands its policy of arming local militias, the ease of these raids may decrease significantly. There have already been examples of local militias interdicting insurgent raids in Muidumbe and Palma districts. As the lean season progresses, this kind of violent non-compliance with insurgent resource gathering raids is likely to proliferate.

Does Ethnicity Drive Civilian Compliance or Support?

Ethnic tension is often the first suspect when interrogating the roots of African conflict. This is also the case in Cabo Delgado, where for years analysts have argued that one of the main drivers of the conflict is friction between ethnic Mwanis and their Makonde and Makua neighbors. Yet the argument that ethnicity is the core cleavage in Cabo Delgado is circumstantial at best. Instead, it is more accurate to understand ethnicity as one element that contributes to the formation of networks that individual insurgents then utilize to gain compliance and support from civilians. It is membership in insurgent networks, not shared demography, that most predicts the extent to which civilians will comply with the insurgency. 

The story of the insurgency as an expression of Mwani grievance comes primarily from two sources of evidence. The first is the historical record. In pre-colonial Mozambique, the mostly Muslim Mwanis held a privileged position in what is now coastal Cabo Delgado due to their slave-trading relationships with larger polities around the Indian Ocean rim. Since independence, however, the largely Christian Makondes have grown in relative power due to their strong support for Frelimo during the independence war. Indeed, Mozambique’s current president Filipe Nyusi is a Makonde from Cabo Delgado. Many Mwanis, in response to rising Makonde power within Frelimo, turned toward Renamo as a political home. Today, distribution of benefits for veterans of the independence and civil wars demonstrates the degree of Makonde identification with the state in Cabo Delgado. Most everyone agrees that Makondes in the province receive these benefits disproportionately due to their close association with Frelimo. 

The other evidence for the insurgency’s supposed ethnic character is quantitative. A well-regarded August 2019 study found that 93% of insurgent attacks to that point had taken place in areas with a low density of Emakua speakers, and a correspondingly high density of Kimwani speakers. That fact, in combination with the insurgency’s then-limited incursions into Makua and Makonde areas on the southern and western edges of the conflict zone, led the author to conclude that “ethnicity is an important factor… in the insurgency.” 

These two explanations, however, tell contradictory stories. If the core cleavage of the conflict were between Mwanis and Makondes, as the historical account seems to suggest, then we would expect an ever-increasing number of attacks to take place in majority-Makonde areas, so that Mwani insurgents could press their grievances. As the quantitative evidence makes clear, however, attacks are taking place primarily in areas where there are many Mwanis. Indeed, since the August 2019 study, the primary areas of insurgent expansion have not been westward toward the Makonde plateau, but south, into the primarily Makua Quissanga district.  

In fact, as the historical account demonstrates, ethnicity is so closely correlated with other important identities — language, religion, national political party, and even profession — that identifying a geographical relationship between ethnicity and attacks does little to explain the driving cleavage underlying the insurgency. A correlation between attacks and ethnicity could just as easily be a correlation between attacks and religion, or attacks and political allegiances. We simply cannot draw many conclusions about the relevance of ethnicity from the geographic data.

We can, however, take clues from the evidence against ethnic essentialism within the insurgency that has come out since the start of the conflict. Researchers have found numerous descriptions of Kimakonde- and Emakua-speaking insurgents (to say nothing of accounts from prisoners of Changana speakers among the insurgents, suggesting members from southern Mozambique). Abdala Likonga, recently identified by the Center for Investigative Journalism as a local insurgent leader, is married to a Makonde woman. Sheikh Sualehe Rafayel, one of the first religious dissidents to preach what would eventually become the theology of the insurgency, is an ethnic Makua, as is Sheikh Abdul Carimo, one of his early followers.

Ethnicity is an important part of identity in Cabo Delgado, and it is one of the major forces structuring the personal networks of those who live there. It is not, however, the only force shaping those networks, or even the primary one. With so little clear evidence of ethnicity as being the organizing factor behind the insurgency, it would be a mistake to see it as the main factor determining levels of civilian compliance with the insurgency. Instead, likelihood of civilian compliance is best thought of as a factor of an individual’s susceptibility to insurgent inducements, both political and financial, and vulnerability to insurgent threats. Religion, politics, and class are all just as powerful predictors of susceptibility and vulnerability as ethnicity.

International and Regional Support

As the insurgency continues, Mozambique has yet to put in place a coherent plan of action around international support; whether it has even developed one remains moot. 

In September, the Portuguese news agency Lusa published details of a leaked letter from Mozambique to the European Union’s High Representative for Foreign Affairs, Josep Borrell, requesting support to train its security forces, and the provision of medical equipment and humanitarian assistance to bolster its efforts to combat the insurgency. European countries, especially France and Portugal, are keen for Brussels to engage, especially in light of the growing threat to LNG investments. The EU has confirmed they are ready to help and that they “have opened a policy dialogue, with a focus on humanitarian, development and security issues in Cabo Delgado.” It is an important opportunity to test and help develop the Mozambican government's longer term strategy beyond the hard security priorities, if Maputo is open to doing so. The EU will understandably tread cautiously in a context of unresolved concerns about the secret debt imbroglio which led to the 2016 suspension of budget support from the EU and other donors. The EU needs to develop better insights into the situation on the ground in Cabo Delgado and the government's plans to utilize any potential support. Civil society groups in Mozambique have questioned why Maputo is turning to the EU, and avoiding domestic debates on key issues relating to the insurgency.  

This move has also highlighted questions about why Maputo is not working more actively with the continental peace and security architecture. Cabo Ligado has emphasized in previous reporting how Mozambique is dragging its heels on submitting its plan of action on requested assistance. It is now almost five months since they undertook to submit a roadmap setting out its needs to the Southern African Development Community’s Organ for Politics, Defence and Security. In the absence of clear communication, this appeal to the EU looks like a slap in the face for SADC, which is especially odd given Mozambique is the current chair. Further, securing SADC support may well have enhanced Maputo’s leverage with Brussels.

Media reports circulating in late September claimed that the US Assistant Secretary for Africa, Tibor Nagy, during a call with Zimbabwe’s foreign minister SB Moyo, requested that Zimbabwe get involved in counterinsurgency efforts in Mozambique, and that Moyo had agreed contingent on the US lifting sanctions on Zimbabwe. Although a number of news agencies ran with this story, Cabo Ligado’s fact checking confirmed that although the issue of Mozambique was indeed discussed during the hour-long chat, no such request was tabled and by extension, no conditionalities raised.

Beyond this story, there have been no public developments regarding possible SADC support during September. President Ramaphosa confirmed that plans were still in the process of being drawn up around the SADC member state report. South African Foreign Minister Naledi Pandor told the parliament’s foreign affairs committee on 2 September that South Africa will respond when needs are more clearly articulated: ‘If it’s more intelligence support, if it’s the SA Navy patrolling the coast, if it is assistance from our Defence Force, we as South Africa stand ready.’ 

What exactly can South Africa offer? The current state of the military raises questions about what is possible. With adequate funding, experts believe South Africa could provide assistance in certain areas, such as coastal security and aerial surveillance. However, there are not enough troops for a field deployment and much of South Africa’s combat and transport equipment is ageing.  

Although there are serious questions about capacity, and there is no formal agreement on assistance, South Africa is not just doing nothing. It has recognized the implications of an emerging Islamist threat in the region. Preparations are quietly ongoing to strengthen intelligence cooperation, as well as around areas of potential bespoke support and deployment (i.e. Special Forces). The Afrikaans weekly Vrye Weekblad reported in late September that a South African arms manufacturer – referred to as the Paramount Group in some social media reports – had signed a deal with Maputo to provide helicopter gunships and armored vehicles. In addition to pilot training, the report refers to specialist dog handling training for deployment in the insurgency.

In her comments, Pandor inferred Mozambique was reluctant to engage with an open hand: “it’s a rather unusual matter because we are dealing with a country which has sovereign rights to its own information.” She emphasized that with Mozambique as chair of SADC and South Africa on SADC’s security organ, as well as chairing the African Union, there was a real chance to strengthen multilateral options for cooperation. 

To what extent is reluctance among regional countries to engage influenced by pre-existing tensions with Mozambique? Veteran analyst Andre Thomashausen suggests that both Malawi and Tanzania may be resistant due to unresolved disputes over historical and contemporary differences around navigating the Chire River (Mozambique refuses to allow Malawi to do so) and a joint exploration zone in the Rovuma Basin (which would benefit Tanzania’s natural gas interests). Such issues may temper perspectives but are unlikely to explain the real reason Mozambique and the SADC have not been able to make tangible progress on possible avenues of assistance. Authorities in Maputo reportedly believe at least one other SADC member is indirectly abetting the insurgency by opting not to clamp down on rebel supply lines through its territory. 

What weight one places on these dynamics remains moot. Both Malawi and Tanzania have large Muslim communities of their own with long historical linkages with Muslim communities on the Swahili coast. Radical elements constitute a very small proportion within these communities. There are strong connections between Tanzanians and the insurgency in Cabo Delgado; there is less detail on Malawian connections. The pro-government Defesa website reports the apprehension of two groups of Malawians, ostensibly recruited for unknown purposes in Mozambique; the first group (of 31) located in the Dzalanyama forests en route to Tete Province and the other (14) near the border with Niassa. The article notes Malawian security forces are working with Mozambican border control officials on such matters. There is no reporting of these incidents in Malawi, and it is unclear whether these cases refer to possible insurgent recruitment or not. Requests for clarification from the Mozambican journalist who broke the story went unanswered.

A primary consideration around any external support for Mozambique in its efforts to counter the insurgency remains the thorny question of who pays. In the current global environment, every country faces an economic squeeze and options for support are understandably limited amongst an array of competing spending priorities. 

Submitting to assistance from a regional force (i.e. SADC and /or AU), if endorsed by the United Nations Security Council, would provide a financing option for security assistance not currently available, as there would be an international obligation to support. It would however mean ceding control over operational security matters to an external force, something Mozambique will be very reluctant to consider. Foreign Minister Veronica Macomo told reporters in late September that this option had not even been considered, as they were focused on their domestic response. This reflects a wider caution within government regarding exploring options for multilateral assistance.