Trends in Political Violence in the Sahel for the First Half of 2020: A Few Comments

The analyst José Luengo-Cabrera periodically posts graphics capturing different trends in violence and displacement in the Sahel; these graphics are indispensable for thinking about conflict in the region, and I really respect his work. He recently posted graphics for the first half of 2020. I want to briefly comment on some of the trends here.

Let’s start with the regional picture:

In addition to the points Luengo-Cabrera makes, here are a few other basic observations:

  • It’s worth repeating often that even though the current wave of crisis in the Sahel began with the 2012 rebellion in northern Mali, most of the intervening years and particularly the last three and a half have been more violent than 2012. Mali is not in a “post-conflict” phase, despite the signing of a peace agreement called the Algiers Accord in 2015.
  • It also bears repeating that northern Mali has, for some time now, not been the most violent zone in the conflict. Kidal, the heartland of the 2012 rebellion, is not even mentioned in Luengo-Cabrera’s breakdown of violent regions. The most violent areas of the current conflict are central Mali (note that Mopti is the most violent region on his list, and that adjacent Ségou is eighth on the list – more violent than Timbuktu) and northern Burkina Faso (note that while eastern Burkina Faso is heavily affected by insecurity and jihadism, it is the north that is substantially more violent).
  • What appears to propel mass violence, in my view, is multi-directional conflict where the key protagonists/decision-makers are not well-known elites. Why is northern Mali less violent than central Mali? Northern Mali has no shortage of militias – but they tend to be led by seasoned politicians and fighters, in some cases by figures who have been political fixtures since the 1990s. In contrast, in central Mali and northern Burkina Faso one finds the violence is often led by people who have emerged as key actors only during the conflict itself, and who were relatively unknown before.
  • The trend lines, particularly for Mali and Burkina Faso, are horrific. In my view much of the increase in violence stems from the compounding effects of previous violence – as I have said before here on the blog, I am skeptical about the idea that COVID-19 on its own triggered major spikes in violence and/or decisively empowered jihadists in the region.

Let’s now turn to country-specific graphics. Here is Luengo-Cabrera’s graphic for Mali:

A few thoughts:

  • The fine print is important here, namely that the fatalities shown for Gao are actually for both Gao and Ménaka; the latter, still-emergent region is obviously part of the tri-border zone that is now the epicenter of the whole Sahel conflict.
  • Note too that within Mopti, the deadliest region, the east (or non-flooded zone) is substantially more violent than the west. Among the factors here may be that according to some Malian experts I’ve talked to, jihadist control is much more consolidated in the west (in cercles/districts such as Tenenkou and Youwarou) than in the east. I think Stathis Kalyvas’ model about contested control driving violence is too schematic (see Laia Balcells’ Rivalry and Revenge, for example, for a more complex view), but this issue of fragmented control certainly seems to be one element in making the east more violent than the west. Additionally, inter-ethnic tensions have repeatedly boiled over into mass violence in eastern Mopti – it is there that the most infamous massacres of the conflict (Ogassagou March 2019, Sobane-Da June 2019, Ogassagou February 2020, etc.) have occurred.
  • Why was 2017 the real turning point to mass violence? Some analysts may immediately answer “JNIM,” referring to Jama’at Nusrat al-Islam wa-l-Muslimin (the Group for Supporting Islam and Muslims, an al-Qaida-sponsored coalition that was announced in March 2017). But the constituent elements of JNIM were all present in the conflict before their formal grouping under that umbrella. Other factors, then, include the spread of the central Malian conflict into eastern Mopti, the emergence of ethnic militias such as Dan Na Ambassagou (which was formed in the final months of 2016), and an escalating cycle of abuses by both the militias and the state security forces (and the jihadists, obviously). This is not an exhaustive list of the forces driving a really complicated conflict, of course. But perhaps in sum one might say that 2017 is the year that various trends really collided to produce an accelerating downward spiral.

Here is Luengo-Cabrera’s graphic for Burkina Faso:

My comments:

  • The puzzle we have in explaining why things really deteriorated in Mali in 2017 is, mutatis mutandis, the same puzzle we have for 2019 in Burkina Faso. Again, one could posit the same basic collision of factors: jihadist violence, inter-ethnic tensions, and security force abuses. A symbol for all of 2019 could be the massacre at Yirgou that opened the year; in that event you have all the elements for multi-directional violence – a (presumed) jihadist assassination, a collective reprisal against an ethnic group, impunity for perpetrators of violence, etc.
  • Another puzzle that I’ve meant to work on is why the Nord region is not more violent. Note that the Sahel Region accounts for over 1,000 fatalities but that the Nord Region has little more than 150. Yet the Nord Region is actually closer to eastern Mopti than is the Sahel Region. One lesson here, then, is that Burkina Faso’s conflicts are not merely a spillover of central Mali’s conflicts.

Here is Luengo-Cabrera’s graphic for Niger:


  • Luengo-Cabrera notes in a follow-on post that it is 66%, rather than 86%, of the fatalities for the first half of 2020 that occurred in Tillabéri. Still, Niger’s trends are fundamentally different than neighboring countries’ because Niger’s deadliest zone used to be far in the southeast, in other words in the zone affected by Boko Haram and its offshoots. 2015 was a bad year in Diffa, as southeastern Niger experienced a wave of attacks, partially representing Boko Haram’s reprisals against Niger for Niger’s participation in the joint Chadian-Nigerien-Nigerian campaign that broke up Boko Haram’s formal territorial enclave in the first several months of 2015. Diffa was already under a state of emergency by February 2015, and has remained under one ever since. In contrast, it was not until March 2017 that the Nigerien authorities declared a state of emergency in parts of Tillabéri and adjacent Tahoua. Things have only worsened since then, and this year looks to be the rough equivalent for Niger of 2017 for Mali and 2019 for Burkina Faso. Meanwhile Diffa is relatively calm compared to the situation there in 2015, or the situation in Tillabéri now.
  • The best thing I’ve read on Tillabéri recently is this Crisis Group report.

Finally, here is Luengo-Cabrera’s graphic for Chad (Mauritania is relatively calm, so I won’t cover it here):

A brief comment is that the areas affected by Boko Haram and its offshoots are deadlier than whatever rebellion(s) are simmering in the north. Daniel Eizenga’s briefing on Chad and Boko Haram from April of this year remains highly relevant for understanding the situation there.

I don’t have much to offer for a conclusion except that things are quite bad, especially in the tri-border zone. I don’t think counterterrorism operations are really helping that much. And in addition to the violence, you have mass and growing displacement (for which Luengo-Cabrera has also made graphics, but I’ll leave that for another time), food insecurity, and many other factors contributing to a really nightmarish picture for millions of people.

Three Recent Journal Articles on Mali and Mauritania

Three journal articles with crucial perspectives on Mali and Mauritania have appeared recently.

1. Adib Bencherif, Aurélie Campana, and Daniel Stockemer, “Lethal Violence in Civil War: Trends and Micro-Dynamics of Violence in the Northern Mali Conflict (2012-2015),” in Studies in Conflict & Terrorism. The abstract:

This article discusses the trends and micro-dynamics of violence in northern Mali. Using a mixed research design, we focus on the violence used by jihadist groups during the first phases of the Malian civil war (2012–2015). Integrating research on civil war and terrorism, we distinguish between direct and remote violence. Quantitative analyses show that the involvement of jihadist groups in this conflict had a strong impact on the level and intensity of violence of all warring parties. Qualitative analysis of data collected during field research done in Mali between 2016 and 2017 complements the quantitative work. It enlightens that relational dynamics strongly influence the decision to resort to violence within non-state armed groups, including jihadist ones, while local contexts often explain temporal and geographic variations in violence.

And one very interesting excerpt, from pp. 12-13 of the .pdf version:

The Timbuktu region illustrates another kind of relational dynamics between secular rebel and jihadists groups. These dynamics affected the level and intensity of violence and show how pragmatic these groups were. The factions of secular rebel groups operating in this region chose not to confront AQIM and fled to areas outside those controlled by jihadist groups. AQIM was the dominant jihadist group in Timbuktu. AQIM took control of Timbuktu on June 28, 2012, and Lere on November 28, 2012, with- out any major violent confrontations. Given their lack of military capability, MNLA leader in the Timbuktu region decided on a pragmatic retreat.93 Furthermore, at that time the relationships between the different communities living around Timbuktu were relatively stable. Armed groups were reluctant, at least in 2012, to exploit latent inter and intra-community tensions to expand their control over these communities as they had done in Gao. For example, the Front de liberation nationale de l’Azawad (FLNA), an Arab militia created in April 2012, was opposed first to the Tuareg rebels and also the jihadists. They occupied a part of Timbuktu on April 26, 2012. Abou Zaïd, one of the leaders of AQIM, asked the FLNA to leave to avoid death among civilians. On April 27, 2012, they left Timbuktu without a clash.97 Among the Timbuktu region, numerous notables had and are going to have relations with jihadist groups. One young Tuareg from Timbuktu and member of the MNLA summarized: “One of the few who rejected any collaboration with the jihadist groups was the colonel Abass who fled with his group to the Mauritanian border [on November 28, 2012]. (…) After that, in Timbuktu, you could find the true terrorists and Arab and Tuareg who were only collaborating with them because of fear or economic interests.”

2. Oumar Ba, “Governing the Souls and Community: Why Do Islamists Destroy World Heritage Sites?Cambridge Review of International Affairs. The abstract:

From Bamiyan to Timbuktu and Palmyra, Islamic fundamentalist groups have willfully destroyed cultural edifices which were listed as world heritage sites. Yet, beyond the criminal acts and their shock value, this article argues that attacks on cultural and religious sites may be viewed as actions embedded in a political project of gouvernement. In this regard, spectacular destruction of cultural heritage may not be simply a signal sent to the international community, but rather an action embedded in a broader political project of governing territory and its inhabitants, aimed at building a new political community based on a new ethos that includes the control of the economyof cultural heritage sites. This article uses the destruction of cultural heritage sites in Timbuktu in 2012 to show the ways in which they fit within the political project of the Ansar Dine jihadist group. Furthermore, the Islamic State’s attacks on cultural sites in Syria and Iraq are also analyzed in light of a political project to govern the territory and communities. The broader implications of this study include the need to pay closer attention to perpetrators’ claims and justifications and to take them seriously, by both international justice scholarship and policy circles. Doing so does not absolve the crimes or mitigate their gravity, but rather allows for better approaches to identify, protect or rebuild cultural heritage in conflict settings.

Ba’s book States of Justice: The Politics of the International Criminal Court has also just come out with Cambridge University Press.

3. Elemine Ould Mohamed Baba and Francisco Freire, “Looters vs. Traitors: The Muqawama (“Resistance”) Narrative, and its Detractors, in Contemporary Mauritania,” African Studies Review (link is to full .pdf version). The abstract:

Since 2012, when broadcasting licenses were granted to various private television and radio stations in Mauritania, the controversy around the Battle of Um Tounsi (and Mauritania’s colonial past more generally) has grown substantially. One of the results of this unprecedented level of media freedom has been the propagation of views defending the Mauritanian resistance (muqawama in Arabic) to French colonization. On the one hand, verbal and written accounts have emerged which paint certain groups and actors as French colonial power sympathizers. At the same time, various online publications have responded by seriously questioning the very existence of a structured resistance to colonization. This article, drawing predominantly on local sources, highlights the importance of this controversy in studying the western Saharan region social model and its contemporary uses.

One interesting paragraph, from p. 261:

To the best of our knowledge, the existing body of academic discussion on this topic—the rivalry between Hassān and Zwāya status groups— predominantly relies on historical documentation. We draw rather on other elements, analyzing, in particular, documents emanating from different media outlets such as newspapers, radio, TV shows, and social media. Our approach is important insofar as it incorporates a significant corpus of nonreferenced bibliographical materials, largely published online on Mauritanian media platforms or newspapers. This methodological option effectively broadens the scope of available sources on contemporary Mauritanian debates and authors. It should also classify and validate such sources as significant elements in the understanding of regional history. Our selection incorporates the authors—often with a limited track record of published materials (books)—who, through their public voices (in Arabic and French), have more clearly delineated the muqawama controversy.

Finally, you can watch the authors discuss the battle of Oum Tounsy in a video here (in French):

Mali: Sharia in Kidal?

My title here is intentionally provocative – the reference is to a recent RFI article discussing new regulations handed down by the Coordination of Azawad Movements (CMA, a bloc of ex-rebels) in Kidal, northeastern Mali, on 30 January. As of 19 February, the CMA was “backpedaling” (see below), but the issue remains a live and contentious one.

From the first RFI article, regarding the initial regulations:

The CMA, which has administered the city for many years, is taking charge of new sectors of security and justice, replacing the State. The rules are stricter: the sale and consumption of alcohol are henceforth forbidden, foreigners* must have a local guardian, and as for the role of Qadi or Islamic judge, it appears strengthened. The inability of the State to assume its responsibilities in northern Mali continues to pose a problem.

The full CMA declaration, signed by CMA President Alghabass ag Intalla, can be found here. Notably, a lot of the press coverage focused on the alcohol ban and restrictions on foreigners, but the declaration also devotes substantial attention to traffic issues and, in particular, says that armed motorcyclists and pedestrians will be brought before the Islamic tribunal and have their bikes and weapons destroyed. There is a a debate to be had over how much any Islamization at work here is    actually subordinate to the CMA’s bid for securitization; it might be going too far to say that the CMA is using Islam as a tool for taking greater physical control of Kidal, but at the very least one can say that Islamization/Qadi-fication is only one part of a larger ambition to expand the CMA’s roles in both security and non-security sectors (including health).

The RFI article caused a bit of controversy because it drew heavily on comments by the researcher Ferdaous Bouhlel, who has been criticized by other Mali specialists (Malian and non-Malian) for allegedly being too close to the CMA. For example:

(Translation: “With researchers like this, the CMA doesn’t need spokesmen any more.”)

My view, however, is that of Guichaoua:

The controversy over Bouhlel, I would say, is a microcosm of two larger debates – (a) have the CMA and Malian Tuareg/Arab rebels systematically obtained more favorable media coverage than they deserve? and (b) is the CMA more nefarious than it sometimes appears in the media?

In any case, there are some other dynamics to highlight here. Recently in one of the courses I’m teaching, civil wars, we discussed Zachariah Mampilly’s Rebel Rulers and Paul Staniland’s “Wartime Political Orders.” To crude simplify things, one point Mampilly makes is that rebels (or ex-rebels?) develop governance models partly through interaction with civilian populations, whose preferences and needs can shape rebels’ decisions. This is what Bouhlel argues – namely, that the CMA is responding to civilian needs for greater security, and that the CMA is drawing on longstanding idioms of governance in the region. One point Staniland makes is that states and rebels (or ex-rebels?) negotiate different arrangements during wartime, including what Staniland terms “spheres of influence.” The CMA and the Malian government are constantly renegotiating their relationship and probing the limits of the other party’s influence (and these are not the only actors in northern Mali or even in Kidal, of course).

Here it’s worth noting that the CMA’s new rules are at least loosely grounded in the 2015 Algiers Accord, which mentions (.pdf, article 46, pp. 12-13) the “reassertion of the value of the role of Cadis [Qadis] in the administration of justice, notably in terms of civil mediation in a way that accounts for cultural, religious, and customary specificities.” Other actors, however, are unpersuaded that the CMA’s rules have any legitimacy – Ahmed Boutache, president of the Committee for Monitoring the Accord (French acronym CSA), denounced the CMA’s rules as “a flagrant violation of the accord…and an infringement of the sovereign prerogatives of the government of the Republic of Mali.” I see these competing statements as not just legal disagreements but also, again, as a way that each side is probing the limits of the other’s authority and legitimacy.

This brings us back to the issue of the CMA’s “backpedaling,” with the CMA’s 19 February statement acknowledging the authority of the Malian state at the local level and expressing willingness for a dialogue over how to move forward on security and the role of the Qadis. Both the CMA and the state, I think, are in essence making offers and counteroffers amid an evolving and unstable situation.

One wishes, meanwhile, that one knew more about who exactly the Qadis were/are/will be:

Personnel, as they say, is policy.

A final point to consider, and one mentioned in the first RFI article linked above, is the issue of influence from Jama’at Nusrat al-Islam wa-l-Muslimin or the Group for Supporting Islam and Muslims (JNIM). RFI quoted an anonymous official from Kidal denouncing the CMA’s new rules as a reflection of JNIM’s pernicious influence. The CMA, which includes some former members of Ansar al-Din, one of JNIM’s constituent parts, is regularly accused of maintaining contacts with JNIM’s leader Iyad ag Ghali. But all of this brings us back to the question of what all these actors want – would ag Ghali be content with a “shari’a-compliant,” autonomous Kidal? Or does he want something more? And was the CMA channeling ag Ghali’s influence – or attempting to undercut it? I’ve tried to get at the complexity of “jihadist politics” in Timbuktu, but there is as much, if not more, to think about in terms of local Kidal dynamics as well.

*I think the CMA is referring to non-Malians here, but I wonder if there is a hint that all outsiders (Malian or non-Malian) could be required to have supervision.

Piece on Jihadism and Politics in Timbuktu for War on the Rocks

This is a belated post to promote an article I wrote last week for War on the Rocks, where I looked at whether the jihadist project has a “political ceiling,” so to speak, in Mali or elsewhere. I took the Timbuktu region as a case study. I also appeared on their “WarCast” (subscription required) to discuss the piece and the broader situation in Mali.

I welcome your comments!

Mali: What Next for the Truth, Justice, and Reconciliation Commission?

RFI has an article on Mali’s Truth, Justice, and Reconciliation Commission (Commission vérité justice et réconciliation, CVJR) that raises some important questions. The CVJR, whose official website can be found here, was created in 2014 with a mandate through 2018. RFI expects that the mandate will be renewed, but at least two key challenges remain:

  1. How can the Commission hear from as many victims as possible? The article mentions that the office in Kidal only opened two weeks ago; even more seriously, victims can face reprisals if they are seen talking to the Commission. Then too there is the problem of severe violence in the center of Mali, particularly Mopti, which creates waves of new victims as well as new difficulties pertaining to victims’ access to the Commission.
  2. How will the Commission’s plans for a victims’ reparations law be squared with plans for a law of “national understanding,” which some critics call an amnesty? (For one commentary on the law, see here, and for one version of the text, see here.)

These are big questions, of course, and debates over “justice” versus “peace” can be extremely fraught. My own thinking on the bigger picture was heavily influenced by Jacob Mundy’s book Imaginative Geographies of Algerian Violence, which deals in part with ways  that forgetting can be just as important to peace as remembering can.

But to move from the big picture back to the details, I was interested to look a bit into the commission’s structure. From the website, one learns that it comprises twenty-five commissioners, directed by a president (Ousmane Oumarou Sidibé, a lawyer and former labor minister) and two vice presidents (former parliamentary deputy Hat ag Baye* and Islamic scholar El Hadj Sidi Konake). One could say that northerners have a large representation on the commission, with the president coming from Timbuktu, one of the vice presidents (ag Baye) coming from Gao, and at least nine of the commissioners having recognizably Arab or Tuareg names. This is not to say that the commission’s balance is off – after all, the north was where the violence began in the current cycle of conflict, and where many of the victims still are. And the other vice president (Konake) is from Mopti, so that region has senior representation too. I guess what is striking is the comparison between this northern-dominated Commission and many other organs of the Malian government, where northern representation is quite thin. On the other hand, one doesn’t want to get too caught up in the politics of representation, which easily becomes an end in and of itself – what matters is the quality of performance.

A final note is that there are several commissioners with connections to Mali’s High Islamic Council, which could mean both that the Commission actively sought out religious leaders as members and/or that the High Islamic Council had a lot of say in who got to sit on the commission.

*Ag Baye replaced Nina Wallet Intallou, who became Minister of Tourism.


Mali: PM Maiga in Timbuktu, and Reinforcements Promised

The violence in northern Mali is made up of multiple interrelated sub-conflicts, which makes the situation there extremely difficult to understand (including for me). I am increasingly interested in trying to better understand the conflict in Timbuktu (city and region), and am working on a longer piece about it. Timbuktu has been the site of some major attacks, including one targeting the United Nations’ Multi-Dimensional Integrated Stabilization Mission in Mali (MINUSMA) in April 2018.

In light of Timbuktu’s importance, I was interested to see that Prime Minister Soumeylou Maïga visited the city on December 14-15. Back in Bamako after the trip, he announced that the government will deploy an additional 350 security personnel to Timbuktu in early 2019 – not just to fight jihadists, but also to try to respond to pervasive banditry (see also here).

Maïga also announced that “the military region of Taoudenit will also be created in 2019,” a reference to the (to my mind, very confusing) plan to carve new regions out of the existing ones. Taoudenit in particular, which used to be part of Timbuktu Region, seems to exist in some kind of quantum state where it is always simultaneously already created and yet to be born. The other day a colleague tried to track down a map of its administrative boundaries, and only found a few rough approximations.

Below are a few tweets from Maïga about the Timbuktu trip. Note that the optics include not just displays of solidarity with the soldiers and displays of the state providing public services, but also public displays of religiosity (in a gesture toward Timbuktu’s religious status).

VOA also has a good report on the trip here.

Notes on the Carter Center’s Second Report on Mali’s Peace Process

The Carter Center is the independent observer designated to follow the implementation of Mali’s peace process as envisioned by the 2015 Algiers Accord. The selection of an independent observer is itself one part of the Accord’s implementation. The Carter Center released its first report in May 2018, and released its second report on 26 October.

Here are my notes on the latter. To me the most striking passages involved (a) the Carter Center’s concerns about the Accord Monitoring Committee (CSA) and (b) the report’s observations about the Operational Coordination Mechanism (MOC) and civilians’ negative perceptions of it in Gao. Here are some key excerpts:

  • The overall tone is mixed, leaning cautiously optimistic. From p. 3: “The observation period was marked by modest but real progress as well as by a significant pause in implementation caused by the presidential election. While progress has been made in disarmament, demobilization, and reintegration (DDR), other obstacles remain, particularly the establishment of the Interim Authorities and the Operational Coordination Mechanism (Mécanisme opérationnel de coordination – MOC) as fully operational. Despite their continued commitment to the agreement, this mixed record underlines the fact that the Malian parties (government of Mali, Coordination des mouvements de l’Azawad [CMA]), and the Plateforme des mouvements du 14 juin d’Alger [Platform]) remain reluctant to advance quickly.”
  • After noting implementation challenges related to the structures created by the Accord and the signatories’ postures, the report goes on to note other challenges to peace. From p. 4: “Two challenges external to the agreement itself impede progress – the crisis in central Mali and criminal economic activity. The crisis in central Mali could overtax the resources initially earmarked for the execution of the agreement, while the ‘criminal economy’ – whose link with the implementation of the agreement has been sufficiently documented by the report of the group of experts established pursuant to United Nations Security Council Resolution 2374 (2017) – slows and discourages implementation.” For background on the crisis in central Mali, this report is a good place to start for Anglophones; for those who read French, I would add this report as well. The report of the UN group of experts can be found here, and my own notes on it are here. Finally, the Carter Center report discusses these two issues (central Mali and criminal economic activity) a bit more on p. 13.
  • The report makes numerous critiques of the Monitoring Committee/Comité de suivi de l’accord (CSA). From p. 6: “Normally scheduled monthly, only three CSA sessions were held during the five-month observation period, due in large part to the presidential election. These sessions lasted only a single day, and sometimes just a few hours. During these sessions, a blockage on a particular topic occasionally led to the suspension or end of a session. The CSA ratifies, often without discussion or formal decision, the actions or agreements made by the parties…The appointment of the minister of social cohesion [see here – AT] is a significant clarification of thegovernment’s presence in the CSA. At the same time, the Independent Observer notes that senior officials of the CMA, based in Kidal, regularly call into question the conclusions or decisions negotiated by representatives in Bamako. The Platform coalition is often marked by wide differences between its members, which impact and slow decision-making.”
  • The report also focuses in on the difference between the formal installation of the interim authorities in northern areas and their actual functioning. From p. 9: “At the regional level, Interim Authorities have been established officially in Kidal (February 2017), Gao and Ménaka (March 2017), and Timbuktu and Taoudéni (April 2017). However, none are in fact operational because they lack budgets to carry out their missions, including the provision of basic services…Over and above these specific obstacles, the Independent Observer expresses concern about the lack of initiative shown by the government to empower the Interim Authorities. Because of the absence of a budget and activities, the Interim Authorities are gradually being undermined and the government’s good faith called into question.”
  • The report has strong words about the MOC, writing that it is operation but deeply hamstrung in Gao, and “not operational” in Timbuktu and Kidal (p. 10). Significantly, the report notes that in Gao, “the population complains of growing insecurity and tends to attribute the increase in banditry and crime to the presence of MOC members.” In other words, the issue is not just about budgets and technical implementation but also about perceptions. The dynamic the report notes is a very dangerous one.



Mali: An AQIM/JNIM Assassination in Timbuktu and Its Aftermath

On 9 September, a commander of the Operational Coordination Mechanism (French acronym MOC) was assassinated in Timbuktu, northern Mali, killed in his car. The commander’s name has been transliterated various ways – Salim Ould M’Begui, Salim Ould Nbekhi, Salim Baghi, and Saloum Ould Becki. From the Arabic spellings that have been given (see here), I would transliterate it Salim Imbighi.

In any case, he was a member of the Coordination of Movements of Azawad (CMA), a coalition of northern Malian armed movements that all played some part in the rebellion of 2012 against the Malian state. The CMA has three major components – the High Council for the Unity of Azawad (HCUA), the Arab Movement of Azawad (French acronym MAA), and the National Movement for the Liberation of Azawad (French acronym MNLA). M’Begui belonged to the MAA and was further, as we will see below, a member of the Awlad Idris/Oulad Idriss, an Arab tribe in northern Mali.

For further background, the MOC – and the patrols it runs – are a key element of the 2015 Algiers Accord, the peace agreement that aimed to prevent a resumption of war in the north following the 2012 rebellion. There are three signatories to the accord: the Malian government, the CMA, and a cluster of pro-government northern militias called the Plateform. The patrols through the MOC are meant to help these diverse groups work together and, by working together, stabilize the north. The Timbuktu MOC was only set up this May, with only around fifty fighters. The MOC there has yet to start its patrols, and the CMA was earlier accused of dragging its feet regarding patrols in both Timbuktu and Kidal.

Mali’s jihadists are, of course, not part of the accord and they have consistently attempted to sabotage the accord generally and the MOC/patrols specifically. This is key background for understanding M’Begui’s murder. On 17 September, Jama’at Nusrat al-Islam wa-l-Muslimin (JNIM) claimed responsibility for the assassination in Timbuktu. JNIM tied the Timbuktu assassination to other assaults on MOCs in the north, including the massive suicide bombing on the Gao MOC in January 2017.

JNIM, a Mali-centric jihadist coalition formed in March 2017, is an official branch of al-Qa’ida in the Islamic Maghreb (AQIM). JNIM’s leadership includes both Malians (its overall leader is Iyad ag Ghali, the infamous Tuareg rebel-turned-jihadist) and non-Malians. JNIM’s largely Malian character contributes to its sophisticated understanding of the political and tribal/ethnic landscape of northern and central Mali. The non-Malian members also have deep experience in the country, though, due to the relationships that some of AQIM’s Saharan commanders and units developed in the years leading up to the 2012 rebellion.

The political dimension of JNIM’s approach helps explain why the claim of responsibility was not a generic public statement but rather a letter to the Awlad Idris. The letter takes pains to soften any outrage on the tribe’s part over the assassination, using three rhetorical techniques:

  1. Religious framing: The letter implicitly asserts that Islam constitutes a common ground of Islam between the tribe and JNIM. More explicitly, the letter argues that M’Begui had apostatized by joining the MOC. The MOC, in JNIM’s framing, targets legitimate “mujahidin” and works with “unbeliever” forces, namely the Malian army, the United Nations’ MINUSMA, and the G5 Sahel’s joint force. The letter presents the assassination as a form of religious justice and even self-defense on the part of the “mujahidin.”
  2. Framing the assassination as a last resort: The letter refers to JNIM’s repeated warnings to “all the sons of the tribes and the Muslims generally” not to join the MOC. The letters also references JNIM’s distribution of “numerous audio, video, and written statements warning about this critical matter.” In other words, the letter suggests that M’Begui had many chances to avoid being killed.
  3. Conveying respect for the tribe: The letter not only addresses the tribe, the author even offers to “arrange a direct meeting” to address any remaining concerns the tribe may have. In general, JNIM is keen to win over northern Malian Muslims (courting “the popular embrace” or al-hadina al-sha’biyya), and some AQIM leaders have long argued that jihadists need to woo the tribes rather than alienating them.

Various commentators noted that the statement was signed not by ag Ghali but by Algerian national and longtime AQIM senior official Yahya Abu al-Hammam. For some commentators, the statement reflected Abu al-Hammam’s ambitions to displace ag Ghali within JNIM (and therefore more an AQIM action than a JNIM one). I’m not sure I would go that far, but it does seem to me that there are various questions to pursue here about (a) internal coalition politics within JNIM and (b) geographical variations in how JNIM operates, not just between northern and central Mali (a theme I explored a bit here), but also within northern Mali. In this case, there are questions to pursue about differences between JNIM’s approach in Timbuktu as compared with its approach in Kidal – although Kidal witnesses its own share of violence, including two even more recent assassinations.

Here it is worth rewinding the tape to 2012-2013, to recall that ag Ghali’s Ansar al-Din (especially the Tuareg politicians who were part of it at that time) was the dominant force in Kidal during the jihadist occupation of northern Mali, while AQIM was most visible in Timbuktu (though it was present elsewhere, and ag Ghali, who was closer to AQIM than some of the other Ansar al-Din leaders at the time, traveled between northern Mali’s different cities and regions throughout that period). Here it is also worth revisiting Rida Lyammouri’s 2016 post “AQIM Never Really Abandoned Timbuktu, Mali,” which includes some interesting detail on Abu al-Hammam and the Awlad Idris. Adam Sandor’s 2017 report for Centre FrancoPaix is also highly relevant here, particularly pp. 16-17. Variations in the jihadist landscape within northern Mali, in other words, are not at all new.

Moreover, we should note that JNIM’s assassination of a CMA leader is a reminder that amid recurring rumors of behind-the-scenes contact between ag Ghali and the HCUA, the two movements – JNIM and CMA – are sometimes violently opposed. The CMA quickly and strongly denounced the murder in Timbuktu and promised to track down the assassins, and the CMA/MAA’s remarks concerning jihadist “infiltration” in Timbuktu sounded none too friendly.

This and other assassinations, finally, are a reminder that northern Malian politics is not just an intra-elite game in which politicians play with other men’s lives, but also a deadly competition for influence and power in which elites’ own lives are very much at stake. As this incident demonstrates, JNIM walks a fine line by assassinating people – on the one hand, it sends a clear message about the costs of working with the MOC, the peace process, and anti-jihadist forces; on the other hand, JNIM risks antagonizing a wide swath of extremely important northern Malian constituencies, and in that way undercutting its own long-term political strategy.

Mali: Another Look at the Presidential Election Results

The “Les Afriques dans le Monde” project at Sciences Po Bordeaux has posted some useful maps and charts on Mali’s presidential elections.

Here are a few takeaways:

  • It’s really striking to see the pie charts that include abstentions. The visuals really underscore the weakness of President Ibrahim Boubacar Keïta’s second term mandate.
  • The post highlights that of more than 65,000 new voters added to the rolls for the 12 August runoff, approximately half of them were in Gao and half in the diaspora. These are the kinds of numbers that have raised eyebrows in Mali.
  • The maps showing vote share by region are also extremely useful. The map of the first round highlights how well IBK did in the north (especially Kidal and Gao) and how poorly he did in Mopti (which also had, far and away, the highest number of polling place closures due to violence. Interestingly, as the authors note, IBK’s main rival Soumaïla Cissé had his best score in Timbuktu (20%), and his second-best in Gao, so this is not a story of Cissé doing well in south and IBK doing well in the north – rather, it’s the story of two candidates with significant northern support amid a divided south, where the share of votes going to other candidates was much higher. Cissé had minimal support in the south, actually.
  • The map of the second round reinforces these patterns. IBK dominated Kidal, but Cissé preserved a substantial vote share in Timbuktu (increasing, actually, to 26% there) and Gao. Only in those two regions, moreover, was the share of people voting greater than the share of people not voting. In the south, again, Cissé had relatively little support. Moreover, abstentions reached 70% in Segou, Bamako, and Sikasso.
  • I would reiterate what I’ve said before, namely that IBK is in some sense not really the president of Mopti (and even, one could argue, Segou). The violence was so severe, and the abstentions so high, that I take the outcome there as a rejection of the process itself.

Mali Elections: Counting Polling Place Closures from the First Round (Regional and Cercle-Level Data)

Hopefully this post doesn’t duplicate work that has already been done elsewhere, but I have not seen anyone go through and give the cercle-level data from the Malian government’s list of polling place closures from the first round of the presidential elections. Let me restate, for clarity: when Malians voted on 29 July, 871 polling places were closed “for various reasons” (mostly connected to insecurity), according to the government. Those closures were concentrated in one region – Mopti, in the center – and even within Mopti they were concentrated in certain administrative cercles, especially the ones that had already been deeply affected by violence.

So here are some figures on the closures (hand-counted by me, so it’s possible that some mistakes crept in, particularly with Mopti cercle and Tenenkou cercle):

Regional Level

Mali has ten regions (formerly eight). The closures affected four of those regions:

  • Mopti: 729
  • Timbuktu: 101
  • Segou: 39
  • Koulikoro: 2

Both of Koulikoro’s closures were in one cercle, Nara, so I won’t go through that in the counts below.

Cercle Level – Mopti

Mopti has eight cercles (see the Wikipedia map here). The closures affected seven of those cercles (the unaffected cercle was Bankass, the southeasternmost cercle):

  • Mopti: 193
  • Tenenkou: 190
  • Douentza: 154
  • Koro: 58
  • Youwarou: 58
  • Djenne: 50
  • Bandiagara: 26

Cercle Level – Timbuktu

Timbuktu has five cercles (unless my information is out of date since the reorganization from eight into ten regions, in which case I hope readers will correct me; in any case, the Wikipedia map is here). The closures affected four of those cercles:

  • Tombouctou (Timbuktu): 47
  • Gourma-Rharous: 40
  • Niafunke: 12
  • Goundam: 2

Cercle Level – Segou (39 Total)

Segou has seven cercles (Wikipedia map here). The closures affected two of those cercles (the two that border some of the worst-hit areas in Mopti)

  • Niono: 32
  • Macina: 7

I may draw some more conclusions from this data in other posts or pieces, but the immediate and obvious conclusion is the Mopti was the worst affected by far, and that the broad north-south band from Niono in Segou up through Mopti (particularly the western half of the region (Tenenkou, Mopti, Youwarou, Djenne, although obviously Douentza was very bad as well) and then into the Timbuktu region was the main corridor of electoral violence in the first round.

The second round figures are here, although I have not yet seen a cercle-level breakdown of closures. But in the second round as in the first, Mopti was far and away the region with the most closures.