5. Why borderlands have become more violent in North and West Africa
Chapter 5 uses a disaggregated database of violent events to show that political violence is more frequent near borders than elsewhere in North and West Africa. Both violent events and fatalities tend to decrease gradually over distance from borders at the regional level. The effect is most pronounced at short distances and roughly 10% of events and fatalities occur within 10 kilometres of a border. The relationship between violence and distance does not vary by the several types of violent events, such as battles or violence against civilians. It does, however, vary significantly over time as discrete episodes of conflict have waxed and waned within the region. Notably, border violence has strongly increased in the last decade: 23% of all violent events are located within 20 kilometres in 2021, against less than 10% in 2011. The chapter shows that the drivers of political violence in borderlands are heavily dependent on the social and political context of each region. The concentration of violence in borderlands is explained by the local strategies of violent extremist organisations, who use these areas to conduct their attacks and mobilise the civilian population, and by the willingness of some states to conduct extra-territorial campaigns against them.
Borderlands are more violent than other regions and political violence tends to decrease with distance from borders in North and West Africa.
Border violence has strongly increased since the early 2010s. Nearly one-fourth of all violent events occurred within 20 kilometres of a border in 2021, twice as much as in 2011.
Local factors explain why some borderlands such as the Lake Chad region or the Liptako Gourma become hotspots of violence while others do not.
Border violence is not solely determined either by state failure or policy or by the strategies of violent extremist organisations. Instead, the interplay of both factors results in border violence.
North and West Africa countries have experienced unprecedented levels of political insecurity in the last decade. A significant proportion of the events and fatalities associated with the rise of violence in the region are located near international boundaries. This chapter explores to what extent borderlands are more violent than other regions, whether the intensity of violence in borderlands has increased over time, and which factors explain why some borderlands have transformed into hotspots of violence. The chapter confirms that violence tends to decrease with distance from borders and is more intense in borderlands in general. The analysis also shows that not all borderlands are becoming more violent. Violence clearly varies across states and regions, which stresses the need to understand the local determinants of conflicts.
There is a clear empirical relationship between incidents of violence and borders in North and West Africa1. Political violence is more frequent near borders than elsewhere in the region and tends to decrease gradually over distance from borders. This suggests that borderlands typically are less controlled political spaces and that armed groups have fewer impediments to movement or other activities within them. More than 4 000 violent events and nearly 19 000 fatalities were observed within 10 kilometres of a border from 1997-2021, the highest incidences in the region. During the same period, 9% of all violent events and 11% of all fatalities fell within the 0-9 kilometre buffer. The next highest percentage of event in any zone occurred in the 10-19 kilometre buffer (8.0%, see Figure 5.1). A peak of violent activities can be observed between 100 and 110 kilometres, with 6% of the events and 7% of the fatalities, likely due to the presence of large urban centres located relatively close to borders, such as Maiduguri in northern Nigeria.
Source: Authors based on ACLED (2021[1]) data. Data available through 30 June 2021. ACLED data is publicly available.
Taken together, 27% of all violent events occurred less than 50 kilometres from a border, 46% occurred within 100 kilometres, and 72% were within 200 kilometres (Map 5.1). This relationship is similar for the number of people killed in these events, suggesting that the lethality of events is not dependent on their distance to borders. The relationship between violence and distance also remains largely invariant when several types of violent events are considered. The proportion of battles, violent attacks against civilians and acts of remote violence reaches its maximum within 10 kilometres of a border and then decreases regularly with distance for all types (Figure 5.2).
Source: Authors based on ACLED (2021[1]) data. Data available through 30 June 2021. ACLED data is publicly available.
Source: Authors based on ACLED (2021[1]) data. Data available through 30 June 2021. ACLED data is publicly available.
The relationship of violent events to borders varies significantly over time as discrete episodes of conflict have waxed and waned within the region (Figure 5.3). For example, the conflicts in the Gulf of Guinea in the late 1990s and early 2000s yielded high percentages of events near borders. During that period, 22% of violent events in Sierra Leone were recorded within 20 kilometres of borders; this rose to 40% in Liberia and 48% in Guinea. The concentration of violent events near borders during this period is explained by the relatively small size of the countries in conflict and by the tendency for armed groups to use border regions as sanctuaries, such as the establishment of the Sierra Leonean Revolutionary United Front (RUF) along the Liberian border in the early 1990s.
Source: Authors based on ACLED (2021[1]) data. Data available through 30 June 2021. ACLED data is publicly available.
The overall percentage of events near borders declined throughout the mid-2000s as conflict changed location, declining in the small states of the Gulf of Guinea and emerging in larger ones like Algeria, Chad and Nigeria. The most recent wave of violence within the region since 2010 is the most alarming since it involves many larger states (Libya, Mali and Nigeria, primarily) while still resulting in an increase in the percentage of events near borders since 2015. This suggests that the number of conflicts or the amount of violence is not just proportional to the size of a country or to the length of its borders.
The overall increase of events near borders in recent years is a troubling trend as it can herald either the relocation of a conflict within a state, the expansion of a conflict across state borders, or both. From this perspective, the period since 2010 has been especially troubling as the percentage of events within 20 kilometres increased every year between 2011 and 2016 to eventually exceed the historical annual average of 23% for 1997-2009 (Figure 5.3). After a brief decline between 2017 and 2018, the percentage of events within 20 kilometres of a border has been on the rise again, reaching 22% through June 2021.
Within the region, the relationship between borders and violent events is not fully consistent over time. While the average annual percentage of events that occurred within 20 kilometres of a border is over 22%, recent years have been as low as 7% (2011) and as high as 26% (2016). This temporal variability is connected to the shifting geography of conflict during these years.
The use of travel times to delineate border regions instead of buffer zones leads to similar conclusions (Figure 5.4). The accessibility model used to determine how far people can travel across the region from any border crossing suggests that the proportion of violent events located near borders was particularly high in the late 1990s, when the civil wars of the Gulf of Guinea made extensive use of border regions. More than half of the violent events observed in the region until 2001 were located within two hours of a border crossing and more than 70% within four hours, which is considered the threshold to define border regions. The proportion of border related events (i.e. within 0-240 min) has experienced a continuous decrease until 2007, where they represented only a third of the total. This all-time low was followed by a second wave of border violence that continues today. In 2019, for example, less than 20% of violent events were located more than four hours away from a border crossing, in regions that are unlikely to be affected by border violence.
Source: Authors based on ACLED (2021[1]) data. ACLED data is publicly available.
The temporal variability of border violence suggests that, as conflict surges in one part of the region, so too does the utility of borderlands to the belligerents. This speaks to the need to consider not just the temporal variability of this regional pattern but its spatial variability as well.
Unsurprisingly, there are significant differences in the relationship between violence and borders across North and West African states (Figure 5.5). Many of the smaller states show a sharp decline in violence as border distance increases. For example, 60% of all violent events recorded in Togo from 1997-2021 occurred within 10 kilometres from a border, which is understandable given that no location in the country is more than 75 kilometres from the nearest border. Other states, both small and large, exhibit this same general relationship but also present significant spikes at large distances from borders, such as the nearly 64% of events found in the 70 kilometre zone in Guinea-Bissau and the 41% of events in the 170 kilometre zone in Mauritania. Such spikes generally reflect the underlying population distribution throughout each state and specifically reflect the presence of national capital cities or other major urban population centres in a buffer zone, including Monrovia in Liberia (25% of events, 100 kilometres) and Abidjan in Côte d’Ivoire (39%, 100 kilometres). This also points to the salience of capital cities and urban centres in political struggles in the region as many movements are seeking to replace or overthrow the state and/or combatting state forces where they are already concentrated.
Source: Authors based on ACLED (2021[1]) data. ACLED data is publicly available.
Taken together, these outcomes support the overall insight that violence tends to decrease with distance from borders while also pointing toward the salience of population distribution as a possible corollary to this truism. In Niger, for example, 41% of all violent events were within 10 kilometres of a border and 53% were within just 20 kilometres. Burkina Faso’s pattern is less stark but still notable as over 23% of violent events were within 20 kilometres of a border. Simply disaggregating the region into states may not be enough to understand the ways in which borderlands function across and within countries, though. For example, within Niger, the western border region has been far more violent than the borderlands shared with Algeria and Libya. This violence is due to groups based in Mali that conduct cross-border attacks against Nigerien and international armed forces operating there, including France and the United States, and to local ISGS (Islamic State in the Greater Sahara) fighters clashing with Zarma and Tuareg communities and self-defence forces. In fact, the Mali-Burkina Faso-Niger tri-border is now at the heart of one of the region’s conflict hotspots in the same way that Liberian borderlands were central to an earlier episode of conflict. In short, while the relationship between borders and violence is detectable within the region, it is highly fluid and mobile over both time and space.
Interestingly, some countries that have experienced the highest levels of violence since 2010 are those that are least likely to reflect this relationship. Figure 5.5 shows how the borders to violence relationship within Libya and Mali is largely invariant by distance, for example. This is undoubtedly due to the population geography of each country, which is not concentrated near its borders in either state. In Nigeria, the absence of a clear relationship between borders and distance to violence is due to the different subnational conflicts around Lake Chad, in the Middle Belt, and in the Delta that have encompassed half of the country for the last decade. In Libya, much of the civil war took place along the Mediterranean coast, where the vast majority of cities and people are located, and not in border regions.
These conclusions based on a series of buffer zones are similar to the results of the accessibility model developed to measure travel times from border crossings (Figure 5.6). This suggests that both approaches can be used to study the temporal evolution of border violence and its variations across countries. The proportion of events located within four hours of a border crossing is naturally extremely high in The Gambia, Togo, Benin, Guinea-Bissau (>95%), due to the size and shape of those countries, and in countries that experience insurgencies in border regions such as Cameroon and Niger (84 and 76% respectively). Cameroon appears particularly affected by border violence with nearly half of the events (49%) located within one hour of a border crossing. Nigeria is more affected by border violence when border regions are based on real accessibility than when fixed buffer zones are used: 72% of violent events occurred within four hours of travel from a border, including 14% within one hour. The density of the road network in Nigeria explain that more distance can be travelled per hour in this country than in the rest of the region, which, in turn, tends to increase the geographical extend of border regions in Nigeria.
Source: Authors based on ACLED (2021[1]) data. ACLED data is publicly available.
Two main epicentres of borderland violence
Border violence is very unevenly distributed across North and West Africa. Violent events and fatalities tend to cluster in specific regions that can become persistent hotspots of violence. These hotspots of violence can be revealed by the Spatial Conflict Dynamics indicator (SCDi) specifically developed to measure the evolution of conflict over space and time (Chapter 3). After dividing the region into 6 540 identical “cells” of 50 by 50 kilometre, the SCDi identifies four types of conflict depending on the intensity and spatial distribution of violence in each cell. Conflicts can be intensifying locally if violence is intense and clustered (type 1), accelerating if violence is intense and dispersed (type 2), beginning or ending if violence is of low intensity and clustered (type 3), or lingering if the intensity and concentration of violence are low (type 4). The application of the SCDi to the region suggests that the four categories of conflict do not appear in the same proportion within borderlands as they do elsewhere: violence is more intense and more clustered in these regions than in the rest of North and West Africa.
Results shows that 30% of the cells of the region are within 100 kilometres of an international boundary and can be categorised as borderland cells. In 2020, 13% of these borderland cells were in conflict. Map 5.2, which shows the SCDi categories for borderland cells that experienced violence, highlights the recent decrease in violence in North Africa and the deepening violence in West Africa. Indeed, north of the Sahara, only eight borderland cells received a SCDi classification in 2020. The remaining 245 borderland cells with SCDi classifications were in West Africa. The map also highlights how crises in neighbouring states can become comingled in borderlands. For example, all of Nigeria’s various sub-national conflicts are expressed in its borderlands, particularly around Lake Chad. The lingering Anglophone insurgency in Cameroon is also affecting the already violent southern Nigerian-Cameroonian border. In this case, two largely disconnected political issues are combining to jointly destabilise the borderlands on either side of the border.
Source: Authors based on ACLED (2021[1]) data. ACLED data is publicly available.
Apart from Nigeria’s troubles, the epicentre of borderland violence in West Africa is along the Malian-Burkina Faso-Niger tri-border region. Along Mali’s eastern boundary, the cluster extends from the Malian border towns of Bénéna in the west to Andéramboukane in the east, a distance of more than 800 kilometres. The cluster also subsumes all the borderlands along the entire Burkina Faso-Niger boundary, a distance of just over 400 kilometres. Taken together, these borderland cells represent a conflict belt approximately 1 200 kilometres long and 200 kilometres wide. This represents a disturbing geographical thickening of violence in the region and speaks to how entrenched violence has become in these borderlands.
Borderlands are more prone to high-intensity violence
A comparison of the types of conflict by region offers some insight into the specificities of border areas. Table 5.1 shows that borderlands are more prone to clustered high-intensity forms of violence than are other regions (Type 1). Between 1997 and 2021, 53% of Type 1 conflicts were observed in border regions, compared with 49% elsewhere. In 2020 alone, this propensity for clustered high-intensity violence was more pronounced as these types of conflicts were observed 59% of the time. The number of regions characterised by a high intensity of dispersed events (Type 2) also increased in 2020.
Taken together, this analysis identifies that, both historically and recently, the proportional mix of SCDi types is notably different in border regions than within non-borderlands. That is to say that when violence occurs in borderlands, it is expressed differently than it is elsewhere and this difference is clearly observable in 2020 (Figure 5.7).
Source: Authors based on ACLED data (2021[1]). ACLED data is publicly available.
The increased frequency of high-intensity violence in general and clustered high-intensity specifically, is worrisome when considering the SCDi’s insights about the life cycle of violence. For example, Type 1 cells tend to represent a conflict in the middle of its life cycle, instead of its beginning or ending (Walther et al., 2021[2]). The fact that borderlands have more of these types of cells than would be expected points to the potential for violence to endure in these regions for some time to come.
The concentration of violent events and fatalities in a border region is partially explained by a state’s capacities and practices towards its borders, which can create historical and geographic contexts under which borderlands can become central to political violence. In border regions where state power is elusive, for example, secessionist movements challenging the authority of the central government can emerge more easily than elsewhere in the country. While borders are crucial to the project of establishing and maintaining political power in general, this study shows that not all borders are always “sites of struggle”. Violence is usually concentrated along certain segments of borders, and at certain points in time, as in the Central Sahel today.
Drawing on these arguments meaningfully puts the focus of explanation on what borders are, what purposes they serve politically, and why one border region might yield more violence than others. This section shows that the drivers of political violence in borderlands are heavily dependent on the social and political context of each region. In addition to a state’s ability to counter external threats, the concentration of violent events along certain segments of borders is explained by the local strategies of non-state actors, who use borderlands to conduct their attacks and mobilise the civilian population. The section focuses on several violent extremist organisations and rebel groups that have successfully exploited border regions to expand their activities, and examines the social and political factors that make these regions a worrying source of political disorder2.
Regional expansion of jihadist armed groups
The regional expansion of jihadist groups in North and West Africa started in the 1990s, amid the Algerian civil war (1991-2002). Algerian jihadists began to cross and exploit borders in various ways. The most prominent Algerian jihadist group of the early to mid-1990s, the Armed Islamic Group (GIA), received delegations from Libya, incorporated some Tunisians, and had a network of rhetorical and material supporters in Europe (Zelin, 2020[3]). Yet the GIA’s brutality, including its murders of some Libyan fighters, prompted backlash from other jihadists around the world, contributing to the movement’s relative isolation by the mid-1990s.
The southward expansion of Algerian jihadists was primarily conducted by the Salafist Group for Preaching and Combat (GSPC), a GIA offshoot more deeply involved in cross-border activity. In the 2000s, GSPC field commanders became central actors in a Saharan kidnapping economy that targeted Westerners in Algeria, Burkina Faso, Mali, Mauritania, Niger and Tunisia. These field commanders included Mokhtar Belmokhtar as well as figures such as Amari Saifi or “El Para,” Abd al-Hamid Abu Zayd (d. 2013), and Yahya Abu al-Hammam (d. 2019). Saifi was captured in 2004 after the GSPC’s first major kidnapping of Westerners, but the others remained major operators in the Sahara and engaged in kidnappings, raids, smuggling activities, and local politics. These forays into the Sahara were built on longstanding patterns of cross-border trade and smuggling (Scheele, 2012[4]).
By the time GSPC rebranded as Al Qaeda in the Islamic Maghreb (AQIM) in 2007, the organisation was well on its way to being a trans-Saharan force. Indeed, its centre of gravity was beginning to shift to Mali rather than its native Algeria. Belmokhtar, Abu Zayd, and others were key players in the jihadist takeover of Mali in 2012-13. That episode thrust AQIM and its offshoot the Movement for Unity and Jihad in West Africa (MUJAO) into the political spotlight in Mali. It also featured jihadists’ own exploratory efforts at border-drawing, as they carved out a short-lived “Islamic Emirate of Azawad.”
Al Qaeda’s move from Algeria to the Sahara-Sahel is clearly visible on Figure 5.8, which represents the proportion of fatalities involving the GSPC, AQIM, and the Group for Supporting Islam and Muslims (JNIM) according to whether they occurred in the country where the group was founded or in another country from 1999-2021. For the sake of simplicity, attacks conducted by GSPC and AQIM in Algeria, and by JNIM and the ISGS in Mali, are considered “domestic”, even though ISGS might be considered equally at home in Burkina Faso, Mali, and Niger. The figure confirms that the proportion of domestic fatalities declined as GSPC became AQIM and extended its operations in the Sahel, from nearly 100% in 2006 to less than 20% in the mid-2010s. In 2017, the formation of JNIM was marked by an increase in domestic attacks – this time in Mali – followed by another regionalisation as JNIM extended its operations into Burkina Faso. The proportion of fatalities involving JNIM outside of Mali exceeds 70% in 2021.
Source: Authors based on ACLED (2021[1]) data. Data available through 30 June 2021. ACLED data is publicly available.
This transnational expansion contrasts with the recent evolution of Boko Haram and the Islamic State West Africa Province (ISWAP), who have remained primarily focused on northern Nigeria, despite an increase in international attacks in the mid-2010s. In 2019, the decision of the Nigerian military to concentrate its forces in fortified camps has led to an increase in the proportion of attacks conducted in Nigeria, which is considered the “home country” of both Boko Haram and ISWAP in this figure. More than 80% of the fatalities involving Boko Haram and ISWAP are located in Nigeria in 2021.
Lack of co-ordinated responses to transnational jihadists
Despite numerous “Sahel strategies” that emphasised the need for a regional response, North and West African states responded to this regionalisation of violence in dispersed order (Walther and Retaillé, 2021[5]). At the regional level, violent events during which military and police forces intervened in another country represent less than 2% of all events in which government forces are involved from 1997-2021 (253 out of 14 049). While all countries have intervened at least once in another country, except for Gambia, Sierra Leone, Ghana and Togo, the fatalities resulting from these events are clustered in a handful of countries: Guinea, Mali, and Nigeria (Map 5.3). In the 1990s, much of the foreign military interventions targeted Sierra Leone (1990-99), Liberia (1997-99) and Guinea-Bissau (1999) under the leadership of the Economic Community of West African States (ECOWAS) Monitoring Group (ECOMOG). In recent years, the vast majority of the victims of foreign military interventions are related to Operation Serval and Barkhane in the Central Sahel since 2014 and the Multinational Joint Task Force (MNJTF) established by Nigeria and its neighbours around Lake Chad since 2015 (OECD/SWAC, 2020[6]). Map 5.3 confirms that most fatalities involving state forces in another country are located in border regions, for example between Liberia and Guinea, Nigeria and Cameroon, or Mali and Burkina Faso. In some instances, a country will seem to deliberately pull away from a strategy of cross-border interventions. Mauritania, for example, conducted raids on Malian territory in 2010 and 2011, then stopped and instead militarised the border zone in a more defensive way.
Source: Authors based on ACLED (2021[1]) data. Data available through 30 June 2021. ACLED data is publicly available.
Violent extremist organisations exploited the lack of co-ordination by expanding or relocating to countries where the political will and military means to counter them was the weakest (Walther and Miles, 2018[7]). In some countries, jihadist organisations benefited from the loosening of border controls, while in other settings their ambitions were curtailed by a militarisation of borders.
North of the Sahara, AQIM benefited from the fall of Muammar Gaddafi in Libya in 2011 following the Arab Spring. Some AQIM commanders such as Belmokhtar, who may have been killed in a French airstrike in Libya in 2016, crossed multiple times into Libya, seeking allies, training opportunities, and profits. AQIM’s presence in Libya, however, did not take on the formalised, territorial character that the Islamic State’s affiliates pursued there, especially during the peak of Islamic State activity in Libya from 2014 to 2016. Indeed, determining who really counted as AQIM could be difficult amid the fluidity of Libyan militias’ and conflict entrepreneurs’ alliances and mutual recriminations (Topol, 2014[8]; Thurston, 2017[9]). Nonetheless, AQIM’s presence remained limited outside of Libya apart from a small AQIM-sponsored terrorist unit in Tunisia, Katibat Uqba bin Nafi, named for an early Arab Muslim conqueror (Zelin, 2020[3]). In Morocco, intensive surveillance, as well as complex policies of religious regulation, appear to have limited the ability of militants to conduct attacks on Moroccan soil, despite a few attacks conducted there by Salafi-jihadists between 2003 and 2011 (Wainscott, 2017[10]), and a relatively large number of Moroccans who travelled to Syria to fight (Sterman and Rosenblatt, 2018[11]).
South of the Sahara, Mauritania initially seemed an attractive target for the GSPC/AQIM. The country was the site of the group’s first major raid outside Algeria, as well as the site of key bombings in 2008, well before major attacks in Mali or Niger began. Yet the amateurish, AQIM-backed local Mauritanian cells faltered amid arrests of their leaders and fighters, and AQIM field commanders began to look for rising opportunities in Mali. Meanwhile, Mauritanian authorities pursued a carrot-and-stick approach, releasing some suspected jihadists under surveillance while detaining the uncompromising hardliners (Thurston, 2020[12]). As part of this strategy, Mauritania heavily militarised its border with Mali and has been relatively free of cross-border jihadist attacks for almost a decade.
Further east, the Mali-Algeria border has remained more open to AQIM, even after the French-led intervention against jihadists in northern Mali in 2013. Since that time, the French and others have prioritised tracking and killing AQIM leaders, and have had some major successes, including killing Abu Zayd in 2013, Abu al-Hammam in 2019, and AQIM’s overall emir, Abdelmalek Droukdel, in 2020. Droukdel was found in part because he dared to travel from Algeria to northern Mali, where he was traced and ambushed by the French. Yet the Mali-Algeria border remains a resource for AQIM and its subsidiary the JNIM, founded in 2017. JNIM’s leader and a long-time partner of AQIM, the Malian national Iyad Ag Ghaly, may spend significant time in southern Algeria, even as some of his lieutenants have been killed in French raids in far northern Mali. Other Saharan jihadists have taken advantage of the border by crossing it to surrender to Algerian authorities (RFI, 2018[13]).
Another key border for JNIM is the Mali-Burkina Faso border. JNIM is a coalition, and much of its expansion has been driven by one of its coalition members, Katibat Macina. Katibat Macina, led by the preacher-turned-jihadist Amadou Kouffa, operates in central Mali and along the Burkina Faso border. An associate of Kouffa’s, the Burkinabè national Ibrahim Dicko, founded a group called Ansaroul Islam (Defenders of Islam) that launched a serious insurgency in northern Burkina Faso in 2016. Although Dicko was killed in 2017, the group lived on, and its partial absorption into JNIM helped the latter become a significant player in the Burkinabè insurgency. The causes of this violence are multiple: localised social tensions in northern Burkina Faso, the heavy-handed security crackdown against the initial violence, the spread of ethnic tensions and civilian vigilantism amid jihadist attacks, and the competition between jihadists and other armed actors to control artisanal gold mining (ICG, 2017[14]). Yet cross-border movement is one element, and various commonalities link central Mali and northern Burkina Faso, including the cross-border presence of the Fulani ethnic group.
The rise of Katibat Macina, Ansaroul Islam, and JNIM have placed the Fulani in a complex position. On the one hand, jihadists have recruited heavily amid their ranks and proclaimed themselves defenders of the Fulani (both Kouffa and Dicko are or were Fulani). On the other hand, other actors’ perceptions of the Fulani as jihadists or crypto-jihadists has led to collective punishment, by state and non-state actors, against Fulani non-combatants (Pflaum, 2021[15]). Real differences between the conflicts in Mali and Burkina Faso persist, and there is significant variation in conflict dynamics from one locality to another even within the same administrative district (Walther et al., 2021[2]). However, there is no question that the ability to win allies and recruits in Burkina Faso has bolstered JNIM’s power as not just a Malian but also a Sahelian entity.
JNIM’s expansion has also raised the possibility that the group will cross other borders, namely the Mali-Senegal border and the borders between southern Mali, southwestern Burkina Faso, and northern Côte d’Ivoire. JNIM units are reportedly present throughout this area, and sporadic attacks on the state security forces in Côte d’Ivoire in 2020 and 2021 have been attributed to JNIM. Côte d’Ivoire is familiar with AQIM and its offshoots — AQIM perpetrated a major terrorist attack at an Ivoirian resort in 2016 — but a sustained JNIM presence in northern Côte d’Ivoire would be a watershed development for the country. Joint Ivoirian-Burkinabè efforts to secure the border initially constrained jihadists’ freedom of movement and operations there (Nsaibia, 2020[16]), although crackdowns can also trigger reprisals from JNIM. The coastal West African states have, on paper, mechanisms such as the Accra Initiative, a co-operation, training, and intelligence-sharing agreement covering Benin, Burkina Faso, Côte d’Ivoire, Ghana and Togo. Yet JNIM’s expansion will test the efficacy of such frameworks.
Transnational jihadism in the Central Sahel
In 2015, ISGS formed as a breakaway faction of Al-Mourabitoun (The Sentinels), itself at that time an estranged faction of AQIM. ISGS’ top leaders have been Adnan Abu Walid al-Sahrawi (killed in 2021) and Abd al-Hakim al-Sahrawi (possibly killed in 2020). In recent years, ISGS has based itself in the tri-border region of Burkina Faso, Mali, and Niger (Le Roux, 2019[17]; OECD/SWAC, 2020[6]). Administratively, ISGS falls under ISWAP, but it appears to be mostly operationally distinct from the Boko Haram offshoot commonly referred to as ISWAP, which operates around the Lake Chad Basin. ISGS operates largely as an extortionist force in a border region, akin to bandits. It simultaneously offers protection to some mobile and/or border communities and perpetrates periodic terrorist attacks on Westerners and other high-profile targets. Today, the borderlands between Mali-Burkina Faso-Niger are one of the most violent regions of North and West Africa (Map 5.4). For instance, since the emergence of ISGS in 2015, nearly half of all the violent events in this region (48%) have occurred within 50 kilometres of the tri-borders.
The presence of ISGS along the tri-border serves to differentiate these borderlands from others in all three of these states. Figure 5.9 compares the percentage of violent events that are within 200 kilometres of the tri-border region with borderlands elsewhere in Mali, Burkina Faso, and Niger. In both categories, violence is highest near a border but there are also important differences. Interestingly, the non-tri-border borderlands display a steep distance decay effect within 20 kilometres of a border and are largely invariant to distance beyond that. The distance effect in the tri-border region is much more gradual with violence occurring between 5 and 10% of all events through 80 kilometres from the border, four times further out than found in the other borderlands. This points to at least two possible typologies of border-distance violence relationships: events that cleave closely to borders and those that permeate borderlands much further away from the border.
The tri-border zone itself has some features that make it particularly amenable to tenacious insurgencies. Many parts of the zone are far from their respective national capitals; the Malian town of Ménaka, for example, is approximately 1 500 kilometres from Bamako, while the eastern Burkinabè flashpoint of Tanwalbougou is approximately 270 kilometres from Ouagadougou. The distances in western Niger are smaller, but can still be significant: it is 200 kilometres from Niamey to the conflict hotspot of Ayorou, for example. In this region as in the rest of West Africa, poor road quality compounds the effects of physical distance (Walther et al., 2020[18]).
ISGS also benefited from pre-existing networks in the region. Prior to the internal split within Al-Mourabitoun, that faction and its leader Mokhtar Belmokhtar had long experience conducting attacks, recruitment, and operations in the Gao Region of Mali and across the border into Niger. Even before the formation of Al-Mourabitoun, Belmokhtar and Abu Walid al-Sahrawi were close to another AQIM splinter group, the MUJAO, which was the dominant jihadist faction in Gao during the 2012-13 jihadist takeover of northern Mali. MUJAO later fused with Belmokhtar’s unit to form Al-Mourabitoun. Al-Sahrawi thus inherited fighters with deep experience in a border region of Mali and with experience conducting attacks in Niger and to a lesser extent in Burkina Faso. At the time that al-Sahrawi pledged allegiance to the Islamic State in 2015, his unit had recently kidnapped a Romanian national at a mining site in Burkina Faso. ISGS found itself first compromising and later clashing with JNIM, who has particular strength in northern Mali, central Mali, and northern Burkina Faso. Accommodation between JNIM and ISGS led to a very loose division of territory that saw ISGS operate mostly to the east of JNIM zones. Even after accommodation between JNIM and ISGS broke down in 2019, ISGS peeled off some JNIM defectors but did not ultimately wrest many territories from JNIM (Nsaibia and Weiss, 2020[19]).
Deep historical patterns of trade, exchange, and pastoralism connect the border zones of Burkina Faso, Mali, and Niger. ISGS thus operates in an area where many of its own recruits, as well as many of its armed civilian enemies as well as surrounding non-combatants, are deeply plugged into cross-border networks and are accustomed to moving across borders. ISGS’ activities, however, have also disrupted older patterns of commerce and have displaced people across borders, for example from Mali into Niger. In Burkina Faso’s Est Region, ISGS found opportunities for political sway and support by controlling artisanal gold mining sites and reopening hunting grounds where the Burkinabè state had restricted residents’ access; these actions reflect broader patterns of Sahelian jihadists targeting gold mines (Lewis and McNeill, 2019[20]).
The tri-border zone is also home to many ethnic groups that are minorities within their respective countries. The Fulani number in the tens of millions, for example, but their numbers are spread out across West and Central Africa, meaning that they account for only 13% of the population of Mali and less than 10% of the population of Burkina Faso and Niger. The Tuareg, similarly, comprise only 11% of Niger’s population and less than 2% of the populations of both Mali and Burkina Faso (CIA, 2021[21]). This dynamic does not automatically mean that governments are hostile to ethnic minorities, but ethnic and linguistic differences can complicate counterinsurgency by erecting communication and cultural barriers between soldiers and civilians, and by reinforcing the tendencies toward ethnic profiling and collective punishment that have marked counterinsurgency efforts throughout the Central Sahel.
Violent extremist organisations such as ISGS tapped into inter-ethnic tensions within the region that themselves reflect competition over land, herding rights, and mobility, including across borders. The most often-cited example relevant to ISGS is the ways in which conflicts between Nigerien Fulani and Malian Tuareg, dating to the 1970s, led some Fulani to seek protection by ISGS (Zandonini, 2019[22]). In Niger’s Tillabéri Region, ISGS recruited among Fulani herders and villagers, offering them protection and enrichment (Bøås, Cissé and Mahamane, 2020[23]). Over time, ISGS’ extortion and predation has elevated inter-communal tensions, which reinforces some recruits’ ties to ISGS but also hardens some ethnic groups’ opposition and resistance to the group, particularly among the Zarma. As the jihadist insurgency spread throughout different parts of the tri-border zone, moreover, the salience of ethnicity increased in many localities, with the Fulani as a whole increasingly stereotyped by state security forces, other ethnic groups and non-state armed groups as terrorists. The ethnicisation of conflict in Tillabéri parallels dynamics that were visible earlier with and around JNIM in central Mali and northern Burkina Faso.
Finally, counterterrorism operations and the mobilisation of ethnically tinged militias helped to push ISGS around within the tri-border zone. Between its formation in 2015 and 2020, ISGS steadily became a greater priority for France, the primary Western security actor in the Sahel. One key turning point was the 2017 ISGS ambush against an American-Nigerien patrol in Tongo Tongo, near Tillabéri. In 2018, France’s Operation Barkhane partnered with two northern Malian militias, the Imghad Tuareg Self-Defense Group and Allies (GATIA) and the Movement for the Salvation of Azawad (MSA) to fight ISGS, especially in the Gao and Ménaka regions of Mali (Nsaibia, 2018[24]). That operation had the unintended consequence of pushing ISGS more into eastern Burkina Faso, which in turn led to the outbreak of a serious insurgency there. By the January 2020 G5 Sahel security summit in the French town of Pau, ISGS was enemy number one for France and the Sahelian states. Yet ISGS’s ability to move within the tri-border zone has made it very difficult to eliminate, even as French strikes have removed many of the group’s leaders.
As the wider Sahel conflict expands, ISGS could spread across other borders as well, namely the Niger-Nigeria border and the Burkina Faso-Benin border. One of ISGS’ most infamous attacks, targeting a group of vacationing French and Nigerien aid workers, occurred in the Kouré giraffe preserve to the southeast of Niamey, not far from the border with Nigeria. There are concerns about an ISGS presence, as well as wider patterns of banditry and Fulani armed communal mobilisation, in southwestern Nigerien border areas such as Dosso and Maradi (ICG, 2021[25]). Given the presence of substantial banditry and organised crime across the border in north-western Nigeria, further ISGS expansion into south-western Niger would have regional ramifications. The presence of ISGS in south-western Niger and eastern Burkina Faso (along with the JNIM presence in the latter as well) also poses a serious threat to northern Benin, where one major kidnapping already occurred in 2019 and where there is a reported presence of several jihadist cells as of 2021 (de Bruijne, 2021[26]).
Transnational jihadism around Lake Chad
The Lake Chad region has become a major epicentre of border violence in the region since the launch of the Boko Haram insurgency in 2009. Boko Haram is a jihadist organisation whose exonym translates as “Western education/culture is forbidden under Islamic law”. Formally known as Jama‘at Ahl al-Sunna li-l-Da‘wa wa-l-Jihad (the Group of Prophet Followers Who Preach and Fight), the organisation was formed in the early 2000s in Nigeria, with the north-eastern city of Maiduguri in Borno State as its base. Maiduguri’s relative geographical location, as well as the evolution of the conflict involving Boko Haram and its offshoots, contributed to the regionalisation of militant violence and recruitment from an early point.
Borno State, as well as neighbouring Yobe and Adamawa states, have long-standing political, economic, cultural, religious, and linguistic ties to nearby areas in the Lake Chad Basin (Hiribarren, 2017[27]). In precolonial times, major polities such as Kanem-Bornu and the Sokoto Caliphate (including its Adamawa Emirate) crossed what are now international borders between Cameroon, Chad, Niger, and Nigeria. Colonial and postcolonial boundary-making has not erased the deep ties connecting ethnic groups such as the Kanuri and the Fulani in the region. While the Kanuri are the dominant ethnic group in Borno and Yobe, and numerous in Niger’s Diffa region, the Fulani are a major ethnic group in Adamawa, in northern Cameroon, and across much of northern Nigeria and the wider Sahel. Boundary-making has also not cut off deep the trade ties connecting Maiduguri to many other markets in the Lake Chad Basin and beyond.
Boko Haram’s initial recruitment base likely drew heavily on in-migrants to the city, and the founders themselves had itinerant biographies (Thurston, 2018[28]). Mobility, including in-country migration but also migration from around the Lake Chad Basin, was thus a core feature of the group from the beginning. The group’s two most important leaders, Muhammad Yusuf (1970-2009) and Abubakar Shekau (d. 2021), were both from Yobe; another pivotal figure, Mamman Nur (d. 2018), was likely Cameroonian. While Nigerian Kanuri from Borno and Yobe have dominated the group’s leadership, Boko Haram has remained open to other nationalities from the region, such as the field commander Mustapha Chadi of Chad (ISWAP, 2018[29]; U.S. Treasury Department, 2015[30]). Boko Haram also attracted recruits from Niger’s Diffa region and other nearby zones prior to 2009, when the organisation launched what became a long-running insurgency against the Nigerian state.
After the 2009 uprising, Boko Haram sought more intensive ties with AQIM and with the Al Qaeda core, although the relationship proved rocky when Shekau bucked external efforts at control over his strategy and operations. Boko Haram’s ties with AQIM involved training, finances, and correspondence, especially between late 2009 and 2011 (Al-Bulaydi, 2017[31]). The presence of Boko Haram fighters was also rumoured during the 2012-13 jihadist occupation of northern Mali, in which AQIM was a key player (Raghavan, 2013[32]). By 2013, Boko Haram’s external jihadist relationships appeared to be slackening, but its mass violence was expanding. This intensification of the conflict came largely in response to heavy-handed Nigerian military operations as well as the rise of the Civilian Joint Task Force (CJTF), a government-backed vigilante force. The CJTF, whose members often had local contacts and intelligence that the Nigerian Army lacked, helped to dramatically curtail Boko Haram’s presence within Maiduguri, but their activities produced a concomitant escalation of the war outside the state capital (Higazi, 2015[33]). During this period, Boko Haram began to attack northern Cameroon as well, transitioning from a previously more passive presence that included recruitment, preaching, and rear-basing dating to 2011 or earlier.
Back in Nigeria, meanwhile, Boko Haram began to capture and overtly hold territory in summer 2014, carving out a “state” that included parts of Borno, Yobe, and Adamawa. Boko Haram’s territorial venture alarmed Nigeria’s neighbours, and Chad and Niger were drawn into greater conflict with Boko Haram in early 2015. A joint Chadian-Nigerien invasion chased Boko Haram out of various towns in north-eastern Nigeria and elicited numerous reprisals, including attacks in Diffa (Niger) and N’Djamena (Chad). Boko Haram units appeared to cross borders with relative ease and developed a substantial presence on the islands of Lake Chad and in other remote zones (Map 5.5). Border towns became recurring flashpoints, such as Gamboru and Ngala on the Nigeria-Cameroon border. In one analysis of data on attacks and clashes, the majority of Boko Haram activity was concerned in Borno and Yobe States, but some units appeared highly active along the Nigeria-Cameroon and Nigeria-Niger borders as well (Prieto Curiel, Walther and O’Clery, 2020[34]).
Source: Authors based on ACLED (2021[1]) data. ACLED data is publicly available.
At the same time, borders somewhat constrained the Lake Chad states’ responses to Boko Haram. Following the Chadian-Nigerien intervention of 2015, efforts to produce an integrated regional response proceeded through the reinvigoration of the Lake Chad Basin Commission (LCBC) created in 1964 and the MNJTF, created in 1994. The MNJTF conducts some patrols and has its headquarters in N’Djamena, although it has mostly been commanded by Nigerian officers. Yet for the most part, the states and militaries of the region continued to act in a parallel rather than fully integrated fashion. For example, Chad’s Operation Boma (or Bohoma) Anger in 2020 received some Nigerian air support but was primarily carried out by the Chadian military (Eizenga, 2020[35]).
In addition to Abubakar Shekau’s organisation, Boko Haram’s key offshoots are the Defenders of Muslims in the Lands of the Blacks (Ansar al-Muslimin or Ansaru), and ISWAP. Ansar al-Muslimin officially launched in 2012 and has been overwhelmingly Nigeria-centric, although it has targeted Westerners and had close links to AQIM. The group had difficulty gaining traction, and some of its early actions led to the arrests of key leaders. ISWAP has been more consequential. During the period from March 2015 to summer 2016, Shekau’s Boko Haram itself bore the name ISWAP after Shekau’s own pledge of allegiance to the Islamic State. In August 2016, however, Islamic State media officially announced that a rival, Abu Mus‘ab al-Barnawi (one of Muhammad Yusuf’s sons) had become the official governor of ISWAP. Shekau, despite continuing to affirm loyalty to the Islamic State, was sidelined and left with a minority of the group’s fighters. Al-Barnawi’s ISWAP eventually killed him in 2021.
ISWAP, like Shekau’s faction, has had a regional presence around Lake Chad, although the heart of its operations remains Nigeria’s Borno State. ISWAP has received fluctuating support from the central Islamic State and may have interacted with the Islamic State’s branches in Libya (Foucher, 2020[36]). At the same time, ISWAP has continued to have heavily Nigerian leadership, especially after Mamman Nur (who had helped engineer the break between al-Barnawi and Shekau) was likely killed in an internal ISWAP dispute in 2018. ISWAP continues to conduct attacks in Nigeria’s neighbours, including the Diffa region of Niger, but has not conclusively expanded beyond the core territories around Lake Chad where Boko Haram has operated since the beginnings of the insurgency.
The Boko Haram crisis also crosses borders through its humanitarian impacts. As of June 2021, internally displaced persons (IDPs) outnumbered refugees in all of the Lake Chad Basin countries except Niger (127 000 refugees and 105 000 IDPs). The epicentre of displacement, both internally and into neighbouring countries, has been north-eastern Nigeria, where there were 2.2 million IDPs as of June 2021 (UNHCR, 2021[37]).
Transnational conflicts in and around Chad
Since Chad became independent from France in 1960, the country has experienced instability on virtually all of its borders (see Box 4.1). In the East, Chad has been affected by the Darfur conflict involving the Sudanese government, the government-backed Janjaweed militias, and several rebel groups (Map 5.6). The conflict in Darfur is often dated to 2003, when rebel groups such as the Justice and Equality Movement (JEM) and the Sudan Liberation Movement (SLM) began sustained attacks, but the conflict has a longer prehistory dating to at least the 1980s, involving tensions over land, water, ethnicity, and politics. Ethnic groups involved in the conflict include Arabs, Fur, and Zaghawa, — all communities that straddle the Sudan-Chad border (Flint and De Waal, 2008[38]). The Zaghawa are the group from which former President Idriss Déby (d. 2021) and many members of his inner circle hail.
Source: Olivier Walther for this publication.
Although Darfur was a key base for Déby during his own rebellion in 1989-90, the war there after 2003 exacerbated various problems for Chad, including cross-border attacks by the Janjaweed, inter-ethnic tensions in the eastern region, and the Sudanese and Chadian governments’ support for rebels seeking to topple the other. Déby was accused of backing the Zaghawa-led JEM, while Sudan’s al-Bashir was accused of supporting the United Front for Change (FUC), the Union of Forces for Democracy and Development (UFDD), and others. Déby faced dramatic rebellions in 2006 and 2008; both times, Darfur-based Chadian rebel groups reached N’Djamena and threatened the survival of the regime. In 2010, Déby and al-Bashir exchanged visits and pledged to stop supporting each other’s enemies, which helped shore up Déby’s power (Debos, 2016[39]). Violence continued in Darfur even after the Chad-Sudan rapprochement, however, with ongoing effects on displacement, farmer-herder violence, and inter-ethnic tensions in eastern Chad. Meanwhile, al-Bashir fell due to popular protests in 2019. The subsequent rise to power of former Janjaweed commander Mohamed Hamdan Dagalo or Hemedti has raised questions about whether Sudan’s new rulers might seek to undermine Déby and his successors (ICG, 2019[40]).
During the 2010s, however, the main base for Chadian rebels became not Darfur but Libya. Déby opposed the 2011 North Atlantic Treaty Organization (NATO)-led intervention against Gaddafi and for the rest of his life blamed that decision for causing chaos in the Sahel. Given the trans-border presence of key ethnic groups such as the Tubu, who have historically supplied some of the leadership of Chadian rebel groups and have been major trans-Saharan traders (Brachet and Scheele, 2019[41]), Libya became a natural destination for northern Chadian dissidents. Haftar, who returned to Libya in 2011 and carved out a de facto polity for his forces in eastern Libya, gave intermittent support to some Chadian mercenaries and rebels (Tossell, 2020[42]). Major rebel incursions occurred in 2019 by the Union of Resistance Forces (URF) and in 2021 by the Front for Change and Concord in Chad (FACT), necessitating French airstrikes against rebels in 2019 and evoking Chadian military deployments both times. Rebels’ survival, planning, and equipping has been aided by their access to Libya as a rear base (Walsh, 2021[43]). In 2018, Chad signed a border control agreement with Niger, Sudan, and Libya’s Government of National Accord; in summer 2021, Mahamat Déby visited Sudan to, among other goals, attempt to revive that framework.
On its southern border, Chad has been accused of backing the Seleka, a rebel coalition from northern Central African Republic that formed in 2012 and overthrew the government of then-President François Bozizé in 2013. The Seleka formally disbanded in 2013 but “ex-Seleka” factions remain key actors in the Central African Republic’s ongoing crisis. Central African Republic rebels and soldiers sometimes cross into Chad, and in 2021, there was a major diplomatic row following Central African Republic soldiers’ attack on a Chadian border post.
The regional relationship between borders and violence is quite clear: borderlands experience high levels of violence and violence tends to decrease as distance from international borders increases. This is true in the region in aggregate despite the fact that the relationship is also highly variable spatially, both among countries and within them. For example, violence is more frequent near borders in some of the states that are currently exhibiting high levels of conflict, such as Burkina Faso and Niger. However, for other countries also currently mired in conflict, such as Mali, Nigeria, and Libya the effect is less pronounced. This points toward the idea that borderlands are not always the spatial epicentres of political conflicts. Indeed, in some circumstances, such as when conflict takes on an urban dimension or is focused upon the control of capitals, violence is observed at much higher rates further from the border than might be suggested by the relationship observed at the regional level.
The study suggests that these border violence patterns can be explained in part by the choices of violent extremist organisations to use borderlands to avoid state forces and to seek recruits from restive or marginalised border populations (see Chapter 2). The relocation of AQIM from Algeria to the Sahel, the opportunistic move of Boko Haram from Borno State to neighbouring states, or the more recent shift from Mali to Burkina Faso of ISGS confirm that transnational armed groups relocate to areas where the government is unwilling and/or unable to counter them (Arsenault and Bacon, 2015[44]). Violent extremist organisations also use borderlands to recruit, train and expand their operations internationally where they can benefit from social networks established in previous waves of the conflict (Walther, Radil and Russell, 2021[45]).
In the Liptako Gourma, for example, ISGS has exploited existing networks and local grievances left unaddressed by the government to present themselves as the defenders of marginalised communities, including pastoral groups. Their expansion was greatly facilitated by the relative absence of state services, confirming the general argument that porous borders contribute to jihadists’ survival and success. In many respects, AQIM and ISGS appear as the most transnational of the various armed extremist groups of the region and the most willing to challenge the informal arrangements established since colonial times between national elites and border populations.
Around Lake Chad, Boko Haram and ISWAP have also shown great ability to cross borders, seek shelter, tap new recruits, and find alternative supply lines that appear linked to the groups’ remarkable tenacity. This was particularly evident in the aftermath of the 2009 Boko Haram uprising, during the movement’s adaptation to the CJTF’s rise in 2013, and after the 2015 Chadian-Nigerien campaign. In other ways, however, Boko Haram and ISWAP have remained deeply parochial movements. Their expansion into border areas of Niger and Cameroon has not been followed by systematic expansion into other parts of those countries, much less into nearby countries such as the Central African Republic. Boko Haram and ISWAP are most comfortable in border zones, rather than in the interior of countries, including Nigeria. Despite some of Boko Haram’s most famous attacks occurring outside of the greater Lake Chad Basin, such as two major bombings in Nigeria’s capital in 2011, the insurgency has always gravitated back to its borderland birthplace.
Given that a presence in borderlands can offer violent extremist groups easy access to either side of a border, states often find themselves in the difficult position of choosing how to counter them. Some governments in the region have shown a willingness to engage in cross-border extra-territorial attacks against extremist groups in neighbouring states. In some cases, government forces operate in formal regional partnerships, such as the ECOMOG, CJTF, MNJTF, or the G5 Sahel, while in others, governments pursue these efforts unilaterally. In both circumstances, such extra-territorial events initiated by governments should be expected to cleave closely to borders for a variety of reasons, such as concerns about straining international norms about state sovereignty or disrupting relationships with neighbouring states. This would limit the impact of such efforts to borderlands and undoubtedly contributes to the higher rates of violence along some borderlands, such as between Burkina Faso, Mali, and Niger, and in the Lake Chad region.
Given that these patterns are the result of the interaction of states and their challengers in the region, it is meaningful to recognise that borderlands are not inherently violent spaces or that borderlanders are somehow destined to be either continually victims or sources of violence. For example, this analysis shows that only at very short distances (20 kilometres or less from a border) do violent events occur at a much higher rate than would be expected. Beyond 20 kilometres, borderlands in aggregate appear similar to non-borderlands and many borderlands exhibit no violence to speak of at all. Further, this analysis shows that the overall regional relationship is also highly variable over time and exhibits an episodic nature as violence ebbs and flows over time.
Taken together, this analysis highlights that nothing about border violence is inevitable and that the current conditions, as dire as they are in many borderlands, can be improved. Governments, international and non-governmental organisations, and civil society groups all have roles to play in deescalating violence and countering the appeal of extremist groups to borderlanders.
References
[1] ACLED (2021), Armed Conflict Location & Event Data Project, https://acleddata.com.
[31] Al-Bulaydi, R. (2017), Nasa’ih wa-Tawjihat Shar‘iyya min al-Shaykh Abi al-Hasan Rashid li-Mujahidi Nayjiriya, http://Jihadology.net.
[44] Arsenault, E. and T. Bacon (2015), “Disaggregating and defeating terrorist safe havens”, Conflict and Terrorism, Vol. 38/2, pp. 85-112.
[23] Bøås, M., A. Cissé and L. Mahamane (2020), “Explaining violence in Tillabéri: Insurgent appropriation of local grievances?”, The International Spectator, Vol. 55/4, pp. 118-132, https://doi.org/10.1080/03932729.2020.1833567.
[41] Brachet, J. and J. Scheele (2019), The Value of Disorder: Autonomy, Prosperity, and Plunder in the Chadian Sahara, Cambridge University Press, Cambridge, https://doi.org/10.1017/9781108566315.
[21] CIA (2021), World Factbook, Skyhorse, New York.
[26] de Bruijne, K. (2021), “Laws of Attraction: Northern Benin and risk of violent extremist spillover”, Clingendael, 10 June, https://www.clingendael.org/publication/northern-benin-and-risk-violent-extremist-spillover.
[39] Debos, M. (2016), Living by the Gun in Chad : Combatants, Impunity and State Formation, Zed Books, New York.
[35] Eizenga, D. (2020), “Chad’s escalating fight against Boko Haram”, Africa Center for Strategic Studies, 20 April, https://africacenter.org/spotlight/chad-escalating-fight-against-boko-haram.
[38] Flint, J. and A. De Waal (2008), Darfur : A New History of a Long War, Zed Books, London.
[36] Foucher, V. (2020), The Islamic State Franchises in Africa: Lessons from Lake Chad, International Crisis Group, 29 October, https://www.crisisgroup.org/africa/west-africa/nigeria/islamic-state-franchises-africa-lessons-lake-chad.
[33] Higazi, A. (2015), “Mobilisation into and against Boko Haram in North-East Nigeria”, in Tall, K., M. Pommerolle and M. Cahen (eds.), Collective Mobilisations in Africa: Enough is Enough!, Brill, https://doi.org/10.1163/9789004300002_014.
[27] Hiribarren, V. (2017), A History of Borno: Trans-Saharan Empire to Failing Nigerian State, Hurst & Company, London.
[25] ICG (2021), “South-western Niger: Preventing a new insurrection”, International Crisis Group, Brussels, 29 April.
[40] ICG (2019), “Avoiding the resurgence of intercommunal violence in Eastern Chad”, International Crisis Group, Brussels, 30 December.
[14] ICG (2017), “The Social Roots of Jihadist Violence in Burkina Faso’s North”, International Crisis Group, Brussels, 12 October.
[29] ISWAP (2018), Khadh‘ al-Waram min al-Khawarij al-Shakawiyya bi-Bay‘at Ahl al-Karam, Islamic State West Africa Province, http://Jihadology.net.
[17] Le Roux, P. (2019), “Exploiting borders in the Sahel: The Islamic State in the Greater Sahara”, Africa Center for Strategic Studies, 10 June, https://africacenter.org/spotlight/exploiting-borders-sahel-islamic-state-in-the-greater-sahara-isgs.
[20] Lewis, D. and R. McNeill (2019), “How jihadists struck gold in Africa’s Sahel”, https://www.reuters.com/investigates/special-report/gold-africa-islamists.
[16] Nsaibia, H. (2020), In Light of the Kafolo Attack, Armed Conflict Location & Event Data Project, 24 August.
[24] Nsaibia, H. (2018), “Targeting of the Islamic State in the Greater Sahara (ISGS)”, Armed Conflict Location & Event Data Project, 21 March.
[19] Nsaibia, H. and C. Weiss (2020), “The end of the Sahelian anomaly: How the global conflict between the Islamic State and al-Qa`ida finally came to West Africa”, CTC Sentinel, Vol. 13/7, pp. 1-14.
[6] OECD/SWAC (2020), The Geography of Conflict in North and West Africa, West African Studies, OECD Publishing, Paris, https://doi.org/10.1787/02181039-en.
[15] Pflaum, M. (2021), “Pastoralist violence in North and West Africa”, West African Papers, OECD Publishing, Paris, https://doi.org/10.1787/63dff519-en.
[34] Prieto Curiel, R., O. Walther and N. O’Clery (2020), “Uncovering the internal structure of Boko Haram through its mobility patterns”, Applied Network Science, Vol. 5/1, pp. 1-23.
[46] Radil, S., I. Irmischer and O. Walther (2021), “Contextualizing the relationship between borderlands and political violence: A dynamic space-time analysis in North and West Africa”, Journal of Borderlands Studies, https://doi.org/10.1080/08865655.2021.1968926.
[32] Raghavan, S. (2013), “Nigerian Islamist militants return from Mali with weapons, skills”, The Washington Post, 31 May.
[13] RFI (2018), “Algérie: reddition du chef terroriste malien Sultan Ould Badi à Tamanrasset”, Radio France Internationale, 13 August, https://www.rfi.fr/fr/afrique/20180813-algerie-reddition-chef-terroriste-malien-sultan-ould-badi-tamanrasset.
[4] Scheele, J. (2012), Smugglers and Saints of the Sahara: Regional Connectivity in the Twentieth Century, Cambridge University Press, Cambridge.
[11] Sterman, D. and N. Rosenblatt (2018), All Jihad is Local. ISIS in North Africa and the Arabian Peninsula, New America, Washington, D.C.
[12] Thurston, A. (2020), Jihadists of North Africa and the Sahel: Local Politics and Rebel Groups, Cambridge University Press, Cambridge, https://doi.org/10.1017/9781108771160.
[28] Thurston, A. (2018), “Political settlements with jihadists in Algeria and the Sahel”, West African Papers, No. 18, OECD Publishing , Paris, https://doi.org/10.1787/0780622a-en.
[47] Thurston, A. (2017), Boko Haram: The History of an African Jihadist Movement, Princeton University Press, Princeton.
[9] Thurston, A. (2017), “Who Counts as al-Qaeda: Lessons from Libya”, Lawfare, 7 May, https://www.lawfareblog.com/who-counts-al-qaeda-lessons-libya.
[8] Topol, S. (2014), “Guns and Poses. Smuggling and subterfuge in the North African desert”, Harper’s Magazine December.
[42] Tossell, J. (2020), Libya’s Haftar and the Fezzan: One year on, 13 January, https://www.clingendael.org/publication/libyas-haftar-and-fezzan-one-year.
[30] U.S. Treasury Department (2015), Treasury Sanctions Senior Boko Haram Leaders, 1 December, https://www.treasury.gov/press-center/press-releases/Pages/jl0290.aspx.
[37] UNHCR (2021), Niger Operational Update, United Nations High Commissioner for Refugees, June.
[10] Wainscott, A. (2017), Bureaucratizing Islam: Morocco and the War on Terror, Cambridge University Press, Cambridge.
[43] Walsh, D. (2021), “Where Did Chad Rebels Prepare for Their Own War? In Libya”, The New York Times, 22 April, https://www.nytimes.com/2021/04/22/world/africa/chad-rebels.html.
[18] Walther, O. et al. (2020), “Mapping travel time to assess accessibility in West Africa: The role of borders, checkpoints and road conditions”, Journal of Transport Geography, Vol. 82, https://doi.org/10.1016/j.jtrangeo.2019.102590.
[7] Walther, O. and Miles (eds.) (2018), African Border Disorders. Addressing Transnational Extremist Organizations, Routledge, New York.
[45] Walther, O., S. Radil and D. Russell (2021), “Mapping the changing structure of conflict networks in North and West Africa”, African Security, Vol. 14/3, pp. 211-238, https://doi.org/10.1080/19392206.2021.1996173.
[2] Walther, O. et al. (2021), “Introducing the Spatial Conflict Dynamics indicator of political violence”, Terrorism and Political Violence, https://doi.org/10.1080/09546553.2021.1957846.
[5] Walther, O. and O. Retaillé (2021), “Mapping the Sahelian space”, in Villalon, L. (ed.), The Oxford Handbook of the African Sahel, Oxford University Press, Oxford.
[22] Zandonini, G. (2019), “Niger, part 3: Guns won’t win the war”, The New Humanitarian, 15 April, https://www.thenewhumanitarian.org/special-report/2019/04/15/niger-part-3-guns-conflict-militancy.
[3] Zelin, A. (2020), Your Sons Are at Your Service: Tunisia’s Missionaries of Jihad, Columbia University Press, New York.