This article concerns governance and violence rates across the ‘ungoverned’ spaces of the African Sahel1. We consider how the dominant narrative for Africa generally, and the Sahel specifically, ‘securitizes’ space, and presents poverty, underdevelopment, and ‘ungoverned’ spaces as security threats to be addressed (Abrahamsen 2005; Keenan 2008). We argue that the terms ‘failed’ and ‘ungoverned’ have become coterminous and common because they benefit various state and international powers within and across the Sahel, who avoid responsibility for the geo-political and economic processes within these spaces. Not only does the term ‘ungoverned’ obscure the actual practices of power within large states with significant under-populated spaces, but it wrongly assumes and accuses those within that space of being more likely to engage in forms of violence that are destabilizing to state structures and external interests. Actual practices of power across the Sahel reveal that large Sahelian states differ significantly in their types of governance, violence rates and trajectories, activities of opposition groups, and long-term prospects for peace.

The concept of ‘ungoverned space’ pervades discussion of global security threats, and dominates analysis and policy approaches to North Africa and the Sahel in particular. The term is intended to refer to both physical territory and non-physical policy space in which there is an absence of effective state sovereignty and control (Piombo 2007; Hazen 2010). In 2003, the Director of the Central Intelligence’s Worldwide Threat Briefing maintained that the threat posed by ‘vast stretches of ungoverned areas – lawless zones, veritable “no man’s lands,”’ demanded ‘a constant level of scrutiny’. At that time, emphasis was placed on the need to focus on ‘ungoverned spaces’ in spite of the fact that such challenges were ‘not occupying space on the front pages’ (Tenet 2003). In the intervening decade, ungoverned spaces have become a staple of the security lexicon of policy-makers, analysts and researchers, and the territories the term describes thereby inextricably linked to terrorism, terrorist ‘safe havens’ and multiple emerging security threats (State Department 2012: 180)

The Sahel region, in particular, has been subject to characterization in these terms, allegedly constituting little more than ‘a scrubby band of ungoverned terrain straddling Saharan and sub-Saharan Africa’ (Schmidle 2009). More recently, the specter of ‘vast ungoverned expanses’ in Mali has featured in US Senate nominations hearings for Commander of U.S. Africa Command (AFRICOM) (Rodriguez 2013), while ‘ungoverned spaces’ also surfaced in the hearing for Director of the Central Intelligence Agency (Brennan 2013). Similarly, in the UK, Prime Minister David Cameron has spoken repeatedly of the need to ‘close down’ ungoverned spaces (Cameron 2013a, 2013b), while a former member of Cameron’s National Security Council has reportedly spoken in terms of how ‘to civilize an “ungoverned space”’ (Forsyth 2013).

‘Ungoverned space’ is presented as a step towards state failure in the Sahel. As Mitchell (2010) notes, the rhetoric of ‘rogue’ states has drastically changed in order to present these areas as the new frontier of terror: “it is no longer just the strong, aggressive and authoritarian states that provoke concern, but also their opposites: those which allow their territories to appear chaotic, cut off, ungoverned or ungovernable”. As summarized by Simon & Tucker (2007) “analysts focus on ‘forests of failure’ rather than trees. Despite catchy terms like ‘the arc of instability’, entire regions don’t collapse; entire states never fail. Not everything falls apart, even when there is no government control.” This alternative perspective on Sahelian spaces and sub-national violence dynamics reveals the deficiencies of theoretical and policy exaggeration in this region.

Overview of Argument

Using data from Sahel and other African states, we show how the argument regarding ‘ungoverned space’ is theoretically thin, based largely on conjecture, and does not reflect the logistical realities and strategies of violent actors. Political violence is about a contest for power, not its vacuum. Large, peripheral spaces may be useful only at a particular stage of conflict, or potentially as a conduit for resource generation. Orchestrating a rebellion from a peripheral area is particularly ineffective if a group plans to make any advancement, generate supporters, or engage with enemy forces. We support our assertions and claims with real-time data from the Armed Conflict Location and Event Data (Raleigh et al. 2010) on political violence from across the Sahel.

Our argument is built on three pillars: we begin with examining theories of spatial governance and specifically question the dominant narrative that suggests links between ungoverned space, state failure and violence. After a review of the main theories, we problematize the concept of ‘ungoverned’ and state capacity interpretations of governance. We suggest that Sahel states are actually effectively governed, albeit by a variety of agents. Given that Sahel states are among the poorest, largest and most underpopulated African states, the reach of state capacity is bound to be limited, and more accurately, directed towards areas of high population and resource wealth (Herbst 2000). The characterization of other spaces as ‘ungoverned’ obscures the practices and exercises of power and governance that take place therein.

The main consequence that we focus on is how the reach of the state creates opportunities for violence. ‘Ungoverned space’ perpetuates a state-centric understanding of governance and conflict, implying in the first instance, that in such spaces the state is entirely absent; and in the second instance, that such absence results in a lack of any effective governance, and therefore conflict. Contrary to the ‘space securization’ narrative, stability and violence are largely products of the geopolitics of groups’, communities’, states’ and international actors’ relationships instead of hinterland size, the politicization of religious and ethnic identities, or environmental considerations.

Our second pillar considers the geopolitics of the Sahel specifically. We argue that the challenges facing the Sahel are less the result of ‘ungoverned’ space than of ‘too many overlapping forms of governance/interests’. The aims and constraints, differences and practices of violent groups, regions, states and international interests are presented in light of contests for power, rather than the absence thereof. We emphasize the agency and agendas of various actors, including biased regimes that may be complicit in the proliferation of violence, and state policies that create instability through neglect, marginalization, corruption and/or collusion.

Finally, we look specifically at the violent actors operating in the Sahel and how they use space. Political violence generally pools in towns, cities and populated areas, but there is a persuasive argument that, during early points in their life cycle, groups take advantage of underpopulated spaces to prepare, strengthen, train, and generate income through illegal smuggling. In particular, the ‘ungoverned’ argument suggests that terrorists specifically benefit from this ‘hinterland’ geography. This is wrong on two counts: the groups that are active in the Sahel are largely local, have national aims, and act accordingly. Branding all organizations as terrorists is a tactic of national governments and international interests who have shared and distinct reasons for seeking to characterize violent actors in these terms. Groups operating in the Sahel are far more likely to be rebel groups dedicated to changing the regime within a state, militia groups organized by regional or local powers to secure power therein, or those engaging in the communal contests. These groups did not seek out ‘underpopulated’ or ‘ungoverned’ territory: most originate from these territories and emerge within their local communities. Like any violent group, they organize from a base of strength first.

Hence, in seeking answers to why groups organize in peripheral, poor, and large regions, it is far more useful to consider peripherality and poverty as reasons for violence organization, over the base logistics of space. Indeed, when the activities of domestic groups are fully considered and understood in terms of their goals and motivations, the immense external focus and concern about the Sahel seems to have been largely generated by the presence of AQIM, a group that, for all intents and purposes, is not especially active. It is perhaps more similar to the roving Lord’s Resistance Army than any feasible global or western threat.

Further, those violent actors (like AQIM) who seek to engage with state or international forces, western interests or others, do not need large, peripheral spaces to do so: violent actors are just as prevalent in areas considered ‘well controlled’ by states. Indeed, this dynamic illustrates a significant tautology – if a group focuses on western interests, they are branded as terrorists and their actions are seen as a function of lacking control, instead of a strategy built to take advantage of environments and institutional structures in which they find themselves.

Theories, Narratives & Realities

A main tenet of the failed/ungoverned space argument is the presumption that states should be internally and externally sovereign, legitimate and capable of delivering public goods, including security, welfare, and development to their citizens and spaces. By this logic, areas excluded from the state ‘net’ are likely to descend into anarchy and violence by allowing non-state violent actors to settle, plan logistics, engage in crime to generate funds, recruit, train and operate therein. Exclusion is typically defined by the presence of active, sovereign control and poor states, such as those found throughout the Sahel, are less likely to have complete state sovereignty. Hence, the types of effective governance being practiced in the poorest states across Africa are the result of choices between the delivery of public goods and the realistic extent of capacity and power.

Governance

There are four main theoretical frames which consider how regimes make choices about space, inclusion and exclusion: Herbst (2000) contends that African states have a favorable or unfavorable political geography, largely determined by state size, population distribution and resource wealth. The type of political geography influences the extent of state capability and control; in turn this regulates how much violence is likely to erupt therein. Small states with even population distributions are expected to be the most peaceful, while large states with underpopulated ‘hinterlands’ are likely to have the most violence, due to the inability of states to effectively police these territories. This is a static approach to governance, predetermined and, seemingly, unaffected by state policies or practices. It is also plainly contradicted by the experience of mass violence in supposedly geographically favorable states, such as Rwanda, Burundi, Eritrea, Liberia, Sierra Leone and others. In many ways, Herbst (2000) provides the theoretical underpinning to the ‘ungoverned space’ argument and uses Sahel state examples extensively in his approach.

A slightly different interpretation comes from Clapham (1986) who argues that different regimes have changing interests, largely based on their ‘legitimizing’ support base. Hence, the geography of governance should be understood as a result of shifting, layered interest: economic interests dominate, to be followed by political relationships, physical abilities etc. In turn, Boone (2003) effectively argues that Western African regimes practice governance based on economic interests and relationships, and this results in four clear categories of governed space: those where the state is directly present and extractive; indirectly present and extractive; indirectly present and non-extractive; and non-incorporated. Finally, Mehler (2004) argues that new institutional configurations of the African states are based on multiple power holders and sources, each connected through alliances, hierarchies and effective relationships to dictate the extent and depth of power across territories.

What each theory effectively argues is that there are limits to central state power: the opportunities and constraints of states in each of these theoretical approaches remain fixed, while the governance process differs in application, spatial presence, and depth. In understanding the realities and limits to power (especially in large states), we accept that regions and groups will be differentially governed, and some marginalized. Those that are excluded tend to live in peripheral areas, have higher rates of poverty and are effectively ‘politically irrelevant’ (Raleigh, 2010). Yet only Herbst (2000) argues that those areas outside of effective ‘central’ and ‘hierarchical’ state power can be considered ‘ungoverned’ due to a vacuum of power.

The present narrative defining the Sahel belt suggests that its large, underpopulated states, with significant Muslim populations, are new hotbeds of terrorism, regional disorder, destabilization, and attacks on western targets. The assumption is that as a significant part of the Sahelian states are ungoverned, unencumbered by control from regimes, armies, etc., violent groups flourish, train, generate support and engage in conflict with local, national, and even international ‘enemies’. In short, its power vacuums are dangerous. However, evidence from African states and Sahel states specifically counter the notion that violence is more often present, or more intense, in areas far from national capitals or underpopulated relative to other areas. At its most basic level, the presumption that Sahel states are amongst the most dangerous on the continent is not quite the case. As Figure 1 shows, the rate of Sahelian violence is not insignificant, but pales in comparison to individual high violence states in Africa. Even considering the drastic rise in 2012, the Sahel remains an actively violent region, but hardly qualifies as a crisis relative to other spaces. Violence for Chad, Mauritania, Mali, Niger and Northern Nigeria combined was similar in intensity to all of DR-Congo’s 2012 violence rate.

Figure 1 

Rate of Conflict in Sahel Versus other Select African States

In addition, the violence in the Sahel is also less fatal than several of the other regions displayed in Figure 1 (see Appendix Figure A1), and fatality rates are largely in line with event occurrences (see Figure 2).

Figure 2 

Rate of Conflict in Sahel with Fatalities

As is clear in both Figures 1 and 2, violence across the Sahel increased at a very limited rate before 2011–2012, and tripled in 2012 from previous rates. These increases within the Sahel are largely due to Northern Nigeria and Mali (see Figure 3). Northern Nigeria has been highly unstable since 2009, as Boko Haram has terrorized civilians and regional governments with increasing frequency and intensity. Regional competition and governance in Nigeria’s northern states largely drive this violence.

Figure 3 

Comparison Across Individual States

Indeed, the overall effects of violence on civilians in the Sahel belt are largely borne by Northern Nigerian residents (Figure 4).

Figure 4 

Violence Against Civilians in Sahel

We can therefore conclude that the risks of violence within and outside the Sahel differ to the patterns that the ‘ungoverned’ space narrative would suggest. The Sahel may be a space for political violence, but its combined violence rate is similar to individual African states. Further, that is largely due to the increase in 2012, which itself is largely attributed to Northern Nigeria, and the active government and civilian attacks of Boko Haram.

Trigger Mechanisms

The ungoverned space narrative is underscored by a context of state failure, and the trigger mechanisms for regional instability are variously presented as poverty, Islam or environmental change. These other narratives of violence in Sahel states are also poor. In detail, the poverty mechanism suggests that ‘disconnected’ areas within states are the hotbeds for modern political violence (Barnett 2004). The Sahel as a poor world region is classified as ‘disconnected’ at best and does experience political violence. But these violence rates shift over time and space. This indicates that spatial characteristics are not the sole concern (or there would be little temporal vacillation). However, these states are no more violent per capita, or over space, than small states2.

An alternative speculation is that climate changes and vulnerability can explain high Sahelian violence rates. Until the overlapping regional crises of 2011–12, most Sahelian violence was couched in environmental crisis terms, suggesting that the negative climate changes in Arid and Semi Arid Lands (ASAL) would lead to direct increases in community based scarcity and conflict. The evidence for this link is shallow: despite fears that climate change may contribute to and exacerbate the present political crisis, conflict rates are not in line with expected trends. For example: there were droughts in the Sahel in both 2010 and 2012. In 2010, overall Sahelian violence rates were not statistically different than the preceding and following years; in 2012, the drought mainly affected Mauritania and Chad, both of which lost over 50 per cent of the crop from 2011 (Oxfam 2012). However, both states remained relatively peaceful in 2012. Mauritania did experience an increase in events, largely due to border events with Mali and intermittent activity clustered in the capital, but these did not concern food, climate, or water.

The ‘environmental crisis’ narrative for the Sahel has been a standard line for decades (as discussed by Swift 1996), bolstered with weak links between rainfall aberrations, presumed land degradation and violence by Herrero (2006); Keita (1998), Bennett (1991); Homer-Dixon (2001) and Baechler (1999) to name a few. Alternatively, other academic work adamantly argues that the risks and vulnerabilities to climate change across the Sahel are a result of marginalization and political exclusion, largely based on group identity and livelihoods. Violence may, or may not, be an outcome of this political process (Turner 2004; Benjaminson 2008; Raleigh 2010b).

Islam is the final mechanism used to suggest how ungoverned space promotes violence. Throughout the recent discussion of the Sahel crisis, a strong current of ‘Islamaphobia’ dominates: Traub (2012) blatantly states that ‘terrorism is only a problem in failed states with significant Muslim populations” and using his own index of failed states as evidence where 13 of the top 20 Foreign Policy Failed states can be defined as large, African and Muslim. Wege (2012) takes up this point, noting that the African continent is characterized by a number of failed states, and Hizbollah (and others) are exploiting these weaknesses. Islamist activists are therefore the ‘primary drivers of terrorism’ across the Sahel; he too, provides no evidence for these points.

The link between large, Muslim, African states and state failure is presented as a transparent assessment by indices measuring the capacity to provide public security, rule of law and basic social services; low levels of democracy and civil liberties; de-legitimization and criminalization of the state; ethnic and elite factionalism; low, unequal economic performance; inability to contain/manage political conflict and the potential presence of large scale insurgency (Menkhaus 2010). On multiple ‘Failed State’ scales, Sahelian states top the list, but the exact positions of states vary considerably from index to index, or even within the same categories of a single index.

A critique of the ‘Failed State’ lists is that they apply ‘broad strokes’ to Africa, and perpetuate a sense of state crisis where it may not exist. Take, for example, the 2012 Failed State list from Foreign Policy: almost all of the Sahel, and indeed most of Africa is at the highest ‘critical’ risk of failure; some of the more stable states in 2012 were Mali (‘in danger’), Algeria, Senegal, Benin, Tanzania, and Gabon (also ‘in danger’). On the Fund for Peace Failure Index, Niger was regarded as in ‘alert’ status in 2012; Chad in ‘serious alert’, Mauritania in ‘Warning’ while Mali in between ‘Warning’ and ‘Stable’. In fact, Chad was regarded as the 4th most unstable state in the world in 2012 despite a clear reduction in conflict, a peaceful election etc. The experience of violence across the Sahel, and the instability within these ‘high risk’ states is at complete odds with their political violence profiles: Niger, Mauritania and Chad have had minimal issues with internal fighting3 as shown by Figure 1.

These examples serve to underscore the general casualness at which countries in the Sahel are treated as ‘in crisis’ or ‘failing’ without much basis in fact. Both the physical size of a country, coupled with Muslim dominance inhabitants, seem to be significant factors in where one lands on a failed state index. Yet, despite painting all states with the same brush, there are sharp variations in how countries have dealt with, and prepared for, internal security threats.

As a whole, the narrative about failed states and mechanisms for increased violence is not based on solid facts or trends in the Sahel. A response to this dominant conjecture argues that the external interpretation of Sahel politics suffers from an aggregation of threats, conditions and states into a distorted, artificially monolithic region of ungoverned Islamic terror. This is an unhelpful fiction (Berschinski 2007) leading to policy goals directed towards strikes on individual ‘terrorists’, and the corrosion of local African support for intervention and change. Simons & Tucker (2007: 390) go further in questioning the original link between terrorism and state failure:

“Contrary to a commonly held view, significant numbers of international terrorists do not come from failed states. Nor do failed states house many organizations that support terrorism. All states consistently fail some portions of their population. In fact, were we to generalize, it should only be along the following lines: from disenfranchised populations can come foot soldiers, from alienated populations can come terrorists.”

Therefore, the argument that ungoverned spaces are violent because they are parts of failed states, Islamist states, or those undergoing some sort of environmental crisis is unsubstantiated. Indeed, the broader narrative about ‘ungoverned’ territory is highly suspect, as is the relationship between underpopulated areas and violence overall.

Hence, instead of the ‘vacuum’ assumed in theories and suppositions about ungoverned space, those that are without a central authority presence are not necessarily spaces of violence. There may, in fact, be violence between state and non-state actors across the boundaries and territories that are differentially governed, but this is not a function of power vacuums, but power contests. Further, according to direct tests of the Herbst thesis, violence rates are not higher in areas with low/indirect/alternative state presence compared to those of effective state presence (gauged by population, garrisons, roads, distance from capital) (see Raleigh and Hegre, 2009& Raleigh 2010a). The most pressing issues in hinterland areas are how non-state actors create forms of governance, and how regions and agents therein deal with the multiplicity of potential ‘governors’ and interests. Indeed, in the Sahel, the areas of most concern are located between regulated and unregulated spaces, and those where non-state agents are actively and effectively competing with the central regime (Peltier 2009).

Who is in Control of the Sahel?

An alternative perspective on 'ungoverned space' is that it does not exist: all spaces that are populated also have some sort and figure of governance. Areas may not be governed in ways that suit powerful interests, but a range of alternatives- whether a traditional authority, communal organizations, rebels etc.- create a system of order4. Contrary to being a prime example of the political fallout in ungoverned space, the Sahel is an effective counter to perceptions of lacking governance. Governance in hinterland regions is often by combinations and permutations of state and non-state local actors. This form of government incorporates some aspects of hierarchy, even if the intent is a nonhierarchical mode of effective governance. In turn, “non-state actors become ‘governors’ in that they systematically engage in rule-making about the provision of collective goods. It would be equally incorrect to assume that the state is absent: central state authorities are present directly and indirectly in areas of limited statehood- through negotiation, contributions, alliances, or direct obstruction, but they lack capacities to centrally and hierarchically govern” (Risse, 2011).

These arguments draws from Campana & Ducol (2011) and Clunan and Trinkunas (2010) who argue much the same, using examples from around the world. For example, it is well established that Sahelian national governments practice a form of indirect rule, using traditional authority leaders and associates. In Niger’s and Mali’s hinterlands, the form of tacit federalism was, Berschinski (2007) argues, to contain the animosity between northern communities and politically dominant, southern communities in the wake of earlier Tuareg revolts. But, he cautions (2007: 96), “such detailed ethnographies are difficult to reduce to talking points. They do, however, produce a picture at odds with a key premise of the ungoverned- space thesis as it relates to West Africa”.

An associated issue with the ‘ungoverned’ space perspective is how alternative groups are situated within the political space. An aggregation of goals and group behavior often results in the characterization of groups as both ‘terrorists’ and ‘foreign’ to the environment in which they operate. Violent groups operating throughout the Sahel have disparate goals, which dictate their level of activity, type of violence, interactions with government forces and overall position within the range of violence occurring therein. They are also reacting to, and motivated by, different aspects of governance. This, more so than their financial and support systems, will dictate how and where they operate (Dowd & Raleigh 2013). Further, regional states’ treatment of and responses to violent groups are distinct, and are neither neutral nor unbiased in how they promote state power. Indeed, the actions of states within domestic or regional arenas is a far more effective explanation for violence rates, patterns and processes than any indication of physical space or ‘absent’ national forces.

Overall, 44.2 per cent of violent events in the region involves state security forces, while the remainder of events are contests among competing militia and rebel groups, violence against civilians, and confrontations with external forces (such as transnational militaries). This almost even split indicates two things: in the first instance, state security forces are far from absent, and are active parties in sustained local, national and regional conflicts. The simplistic assertion that ‘ungoverned’ spaces are territories of government absence is simply and directly refuted by the high degree of state involvement in conflict activity in these supposedly lawless zones. In the second instance, while national forces are certainly a target for militant groups, contests for local power between alternative sources of security, governance and authority in the form of rebels, militias and communal groups also continue to shape the regional conflict profile of supposedly ‘ungoverned’ spaces in a significant way. The second tenet of the ‘ungoverned space’ thesis – that the absence of physical state presence and authority implies an absence of governance mechanisms of other kinds – is also refuted by these data.

An actor-based analysis also presents a challenge to the narrative of ‘ungoverned spaces’ as it is typically articulated. In the past two years, much of the conflict in the Sahel has been attributed to the presence of al-Qaeda affiliates, although AQIM is the only formal al-Qaeda affiliate active in the region, and the group’s activities actually represent only 3.5 per cent of all non-state violence in the Sahel since 1997, and result in just under 2 per cent of recorded fatalities.5 For several years, AQIM and its predecessor, the Salafist Group for Call and Combat (GSPC), tried to integrate into northern communities, only to be dismissed by native populations as foreign. Simon & Tucker (2007) explain this as a process as ‘ethnic inaccessibility’ and it is found throughout underdeveloped areas: ethnic inaccessibility is compounded when the community is remotely located or difficult to access and/or when groups have been treated as though they deserve to be marginalized. Throughout much of the world, and especially in areas with no sustained physical government presence, terrorists can only successfully hide or train if they secure local support and/or local silence; in such locales, groups need to be able to count either on communal ties and codes of honor and/or on sympathetic elements within the security services. In face of this opposition, over time, AQIM has sought to establish ties with local communities in the Sahel through a deliberate process of social and economic integration (Goïta 2011), the necessity of which itself refutes claims that the group is highly mobile and fluid in ‘ungoverned’ territory in Northern Mali.

Further, the only significant inroads the AQIM and other transnational groups made across the Sahel came after local agents agreed to strategic alliances. Those alliances were loose, as seen in the unstable association between Ansar el Din militants and AQIM, and indicate that most of these groups are not international terrorists as defined by governments and western interests, but local militant groups who are organized against national regimes. They share this feature in common with other violent groups active in regions beyond the Sahel. In particular, AQIM faced several problems in Northern Mali, as Tuareg and other ethno-nationalist groups initially fought against them. The turn in fortunes for AQIM in the region came only after Tuareg forces were prevented from actively fighting by the Malian Military Forces, who did not have the means to do so. Burbank (2010) finds that AQIM forces grew fourfold in the past years due to few local enemies.

To represent Sahelian violence as generated and dominated by foreign (i.e. Algerian) rebels with a ‘global’ jihad agenda is to mischaracterize the region, the conflicts and the crises that occurred in both in 2012 and over the longer term. AQIM are but one group in the entire region; the number of discrete violent groups vacillated from 23 non-state violent groups active in 2007 to 27 in 2010, and 13 discrete groups in 2012.6 These groups are largely domestic, and fighting for national goals and political change. The role of regional groups, with ties to groups such as Al Qaeda is also often directed against, or colluding with, Algeria and while attacks on Western targets and interests are the most high-profile events, these are actually extremely rare (this is discussed at length in Dowd & Raleigh 2013). Attacks on international civilians and aid workers make up only 3.8 per cent of all recorded incidents of violence against civilians. Military forces are active in states that made concerted efforts to limit violence within hinterlands, and those that did so were largely successful, despite the significant space to operate. Table 1 offers a description of each active politically violent group in the Sahel, their goals and patterns of behavior.

The most active area in the Sahel is northern Nigeria, which is home to the most active group- Boko Haram- and a number of violent communal and ethnic militias. Additionally, among other prominent actors over the course of the dataset are the Union of Forces for Democracy and Development (UFDD) in Chad, the National Movement for the Liberation of Azawad (MNLA) in Mali, and the Movement for Democracy and Justice (MDJT) in Chad, and the Nigerien Movement for Justice (MNJ).

That these groups are native to the Sahel region and act on explicitly articulated (ethno-) national agendas counters the implicit assumption that ‘ungoverned spaces’ pose a danger because they draw to them highly mobile, transnational militant actors. These groups emerge in local contexts in response to specific, often communal issues, largely centering around the highly political issues of perceived marginalization; the allocation of national resources and wealth; and claims to autonomy. The shortcomings of this paradigm are particularly clear in the Malian case: under the 2006 Algiers Accords peace agreement, the Malian government committed to the economic development of its north-eastern region, greater sub-national autonomy, and the creation of special Tuareg security units to police the area. The failure to adhere to these conditions led in turn to deteriorating security and the collapse of the peace process in a reinforcing cycle of conflict, underdevelopment and marginalization. The irony is that while analysis of the present Malian context is couched in terms of ‘ungoverned space’ and the need to bolster and consolidate state sovereignty throughout the territory, among the main parties opposing the government are armed groups which sought precisely to establish alternative structures of governance, authority and security provision.

These actors and other ethno-nationally oriented groups throughout the region are not rootless militants opportunistically capitalizing on the chaos of ‘ungoverned’ or ‘lawless’ spaces. Nor is it necessarily the case that the appropriate security and policy response to these challenges should involve state-building, bolstering central authority, and the explicit depoliticization of challenges to the state by framing them as functions of insufficient control or state presence in a region, when deliberate state policy and strategy is being contested.

Roaming areas

Our final point relates to the use of space by groups, both domestic and foreign, throughout the Sahel. For each point in a revolution or rebellion, the type and use of space is dictated by both political and physical considerations. McColl (1969) observes that the use of under-populated space is beneficial at an early stage of mobile war, where the intent is to gather strength, determine logistics, secure financing etc. The strength of the group relative to that of the government(s) will dictate both how long an initial stage lasts, and how close to major cities a group’s designated ‘area’ can be located. The conditions that allow for particular regions to be more hospitable than others include a previous experience in revolution; access to important military and political objectives (including provincial capitals); areas of multiple, confused, overlapping authorities, either by international border or by multiple local power centers; economically self-sufficient, and suitable for military training. Overall, stability on the local or national level should be lacking. Naturally, all bases will not meet all these criteria, but the most important consideration is that the area allows for military engagements and also space for hit and run attacks (McColl 1969). After this mobile stage, the group should progress into the ‘creation of core areas’ wherein an insurgent movement is established, and provides the daily necessities of populations therein. Different to the mobile stage, a core stage is a point of concentration, which allows for standard engagements with military forces. Groups that cannot achieve this stage, either due to their own weakness, or that of the government, are unlikely to have a stable or successful transition to power.

McColl’s (1969) bases for revolutions are helpful in this discussion as it presents an alternative rationale for why some groups may organize in underpopulated regions, and the political benefits and consequences therein; it also provides a further counter to the notion that an initial stage ‘roaming area’ for groups is an ultimate goal, capable to unsettle entire regions. Indeed, a perpetual state of ‘mobile war’ and hit-and-run attacks is much more likely to be a sign of weakness over strength. Bosi (2013:80) takes up these points by arguing “the opportunities provided by safe territories are not necessarily conducive to the continuation of political violence, although they facilitate its persistence over a long period of time”.

This argument is particularly relevant to the Sahel, as its space, demography, overlapping and multiple authorities etc., provide an ideal stage for ‘roaming’ and/or ‘mobile’ war. But this largely depends on the goals of the movement and their innate strengths. As discussed above, most of the active violent groups within the Sahel are local to the region, and their use of space therein is largely confined to their ethno-regional area. This is evident even for those native groups that bear an “islamist’ mantle, such as Ansar El Din. Overall, the use of space by these groups within these territories is largely in line with typical civil war and militia patterns: larger attacks are directed towards towns and cites in the region (e.g. Gao, Timbuktu and Kidal in Mali) in order to maximize the benefit of fighting and holding territory. In turn, attacks on civilians are more widespread, often involving local grievances that are subsumed under the conflict’s dominant cleavages (see Kalyvas 2006). Hence, the vast majority of the active agents within the Sahel use this space to engage in contests over local and regional goals that motivated their initial formation, and not to coordinate, prepare and plan for larger attacks outside the region and the country. The scale of the group goals should be considered in light of their strengths and abilities to coordinate political violence outside of the locality. See Figure 5 for a review of active spaces.

Figure 5 

Overlapping Territories of Action in the Sahel

Yet, there exist a small number of groups who, although native to the larger region, have relocated to neighboring spaces and localities to take advantage of space, smuggling routes or distance from Algerian security services. AQIM and MUJAO7 are examples of groups whose minimal activity, but significant financing, has supported local allied groups in the area to engage in overtly ‘islamist’ violence, and attacks on sites important to the larger region and external parties (e.g. Timbuktu libraries, In Amenas facility). We argue that the use of space by these groups in largely in keeping with an initial stage of revolution that AQIM and its affiliates are hard pressed to graduate from. Large amounts of space, within which groups can act without recourse, is beneficial to groups who are largely coordinating against domestic governments and do not have significant abilities to counter formal forces, despite preparing for years. Their overall threat is diffuse, but not likely to create a more substantial risk for multiple reasons: 1) they lack local community support (in part because the goals are not localized); 2) their capabilities are very much within the realm of hit and run attacks or violence against civilians; and 3) they are engaged in internal fissures, and an inability to broaden their appeal and message. Indeed, their inability to take on significant force was clearly shown in the 2012 attack by Malian, French and Allied African forces in northern Mali. In several cases, the combined, French led, forces did not engage with any opposition as they quickly made their way through the most ‘embattled’ territory. While this may serve as a testament to the reputation of French troops, it is perhaps more likely to be due to the poor condition of AQIM, Ansar El Dine, MUJAO and other allied groups. Their threat is far more likely to be felt in cases like In Amenas, which has been suggested as continued acts of desperation stemming from internal rivalry in AQIM (Burbank, 2010). These occasions, while very unfortunate, will not ‘unsettle’ entire regions, but are rare events designed to position AQIM within a global spotlight, and reinforce the Sahel ‘crisis’ narrative.

The logic behind AQIM activity in the region also suggests a poorly coordinated and ill structured group: the infrastructure and logistical problems in the Sahel, and other under populated regions are significant drawbacks and disincentives for anyone seeking a reliable base for operations (Simon & Tucker 2007: 389). If these minimally active groups could do more than roam, while attacking unarmed civilians and random Algerian targets, the area would have seen considerably more conflict. It is more likely that AQIM is as real and deep a threat at the Lord’s Resistance Army presently active in Central African region. They, like AQIM, left their home area not out of strength, but out of weakness. They survive by attacking marginalized, poor, peripheral communities while claiming an ongoing war with a military power they never engage with. Similar to the campaign to rid Central Africa of the LRA discounted the perpetual and very real threats of other Congolese rebel groups, the Sahelian case is reminiscent of a ‘rebel with no clothes’.

Conclusion

In brief conclusion, this article argues that a range of actors extensively and effectively governs ‘ungoverned’ spaces. Further, in treating the Sahel as a monolith, analysts and those espousing a ‘crisis’ narrative misrepresent the variation in violence, the politics and governance practices of states, the risks of instability, and the political actors therein. Finally, the groups operating within the Sahel are, by and large, domestic groups operating within their own subnational contexts, and challenging local, regional and national governance over issues of corruption, marginalization, political exclusion and mismanagement. The overt focus on the few regionally foreign actors, who perpetuate a small proportion of the political violence, obscures the realities of Sahelian violence. In turn, this serves an ‘external’ interpretation of terrorism and instability largely driven by national governments seeking to re-allocate blame for poor performance and a plethora of active armed actors.

Our argument presented here is relevant to policy and humanitarian decisions as it advocates two new perspectives on Sahel state violence. The first is that political violence within states can only be understood in terms of the capacity, quality, reach and character of governments. Analyses of the Sahel which presumes ‘an inability to govern’ and state failure as starting points obscures these features of political power, and leads directly to arguments which priorities discussions of global terrorism in the Sahel, at the expense of more contextualized understandings and solutions. The second is that the trigger mechanisms often advocated in academic and popular narratives of Sahelian violence obscure the daily political dynamics that shape civilian violence risks, program effectiveness and the adoption of long term stability solution. Public analysis connecting Africa’s poverty, large Muslim populations, and weak governments to the rise of loosely defined ‘terrorism’ exaggerate, distort and ignore the continent’s pressing concerns while leading to policy outcomes of negligible worth (Berschinski, 2007). These discourses serve the interests of the international community, who often ‘aggregate’ threats to simplify the politics in complex area, justify interventions, and engage in regime change.