Abstract

Contrary to common assumption, major forms of large-scale organized political violence in sub-Saharan Africa are declining in frequency and intensity, and the region is not uniquely prone to the onset of warfare. African civil wars in the late 2000s were about half as common compared to the mid-1990s. The character of warfare has also changed. Contemporary wars are typically small-scale, fought on state peripheries and sometimes across multiple states, and involve factionalized insurgents who typically cannot hold significant territory or capture state capitals. Episodes of large-scale mass killing of civilians are also on the decline. That said, other forms of political violence that receive less attention in the academic literature are increasing or persistent. These include electoral violence and violence over access to livelihood resources, such as land and water. While primarily descriptive, the article posits that geo-political shifts since the end of the Cold War are a leading candidate to explain the changing frequency and character of warfare in sub-Saharan Africa.

The fiftieth anniversaries of many sub-Saharan African states provide a critical opportunity to evaluate the post-colonial experience in a number of dimensions.1 Given the coercive nature of colonial states, the divisive colonial boundary-making processes, and the often abrupt transitions to independence, a central experience for many African states has been large-scale political violence. Since independence, many states have witnessed civil wars, mass killings, and other forms of direct political violence. Indeed, so strong is the association between war and independent Africa that many journalistic observers claim that Africa is the most endemically violent region in the world.2 A recent account referred to messy, ‘forever’ wars that ‘never end’ and are ‘spreading across Africa like a viral pandemic’.3 With much greater nuance, scholarly accounts also present a picture of a continent where large-scale violence is common.4

Yet fiftieth anniversaries also provide an opportunity to evaluate critically assumptions about the endemic quality of political violence in Africa. To that end, this article asks four main questions. First, what have been the major longitudinal trends of warfare and large-scale killing of civilians in Africa? Second, is Africa different from other regions in terms of the prevalence and duration of these major forms of political violence? Is Africa in fact the world leader in such violence across time? Third, how is the character of warfare and other forms of political violence changing in sub-Saharan Africa? Finally, what might explain the changing patterns of violence, in particular with regard to armed conflict? Taken together, these questions allow for macro-level observations about different forms of violence that are not usually considered simultaneously and to situate these observations in a comparative regional perspective.

The principal finding is that in the twenty-first century both the volume and the character of civil wars have changed in significant ways.5 Civil wars are and have been the dominant form of warfare in Africa, but they have declined steeply in recent years, so that today there are half as many as in the 1990s. This change tracks global patterns of decline in warfare.6 While some students of African armed conflicts, such as Paul Williams, note the recent trend,7 it is fair to say that the change in the prevalence of civil wars is not recognized by most Africanists and generalists. Equally important but even less noted is that the character of warfare in Africa has changed. Today's wars are typically fought on the peripheries of states, and insurgents tend to be militarily weak and factionalized. The large wars that pitted major fighting forces against each other, in which insurgents threatened to capture a capital or to have enough power to secede, and in which insurgents held significant territory – from the Biafra secessionists in Nigeria, to UNITA in Angola, RENAMO in Mozambique, the TPLF in Ethiopia, the EPLF in Eritrea, the SPLM in Sudan, the NRM in Uganda and the RPF in Rwanda – are few and far between in contemporary sub-Saharan Africa. Somalia's Al-Shabab holds territory and represents a significant threat to the Somali federal transitional government, but given the 20-year void at the centre of Somalia the case is not representative. In April 2011, rebel forces in Côte d'Ivoire captured Abidjan, but they did so with external help and after incumbent Laurent Gbagbo, facing a phalanx of domestic, regional, and international opposition, tried to steal an election.8

More characteristic of the late 2000s and the early 2010s are the low-level insurgencies in Casamance (Senegal), the Ogaden (Ethiopia), the Caprivi strip (Namibia), northern Uganda (the Lord's Resistance Army), Cabinda (Angola), Nigeria (Boko Haram), Chad and the Central African Republic (various armed groups in the east), Sudan (Darfur), and South Sudan, as well as the insurgent-bandits in eastern Congo (a variety of armed actors, including Rwandan insurgents) and northern Mali (al-Qaeda in the Maghreb). Although these armed groups are in some cases capable of sowing terror and disruption, they tend to be small in size, internally divided, poorly structured and trained, and without access to heavy weapons.9 Several of today's rebel groups have strong transnational characteristics, that is, insurgents move fluidly between states. Few are at present a significant military threat to the governments they face or in a position to seize and hold large swaths of territory.

The change in quantity and character of civil war presents an analytical puzzle that is not well explained by the existing academic literature. William Reno's insightful book on the history of African insurgencies demonstrates the shifts from anti-colonial, to Marxist-Leninist, to ‘reform’ (such as Yoweri Museveni's NRM), to warlord (such as Charles Taylor's NPFL) rebels. His central thesis is that insurgent strategies and tactics reflect the nature of political authority in the states in which they fight.10 But today's weak and factionalized insurgents are not simply reflections of the political authority in their states. Most contemporary insurgencies fight on the periphery of fairly well consolidated states, as in Senegal, Nigeria, Angola, Namibia, Mali, Sudan, and Uganda.11 The dominant political science theories of civil war emphasize slow-moving (or immutable) conditions. Such factors include weak states, rough terrain, natural resource (typically mineral) dependence, and horizontal inequality.12 But these variables do not account well for the sharp decline and change in character of warfare in sub-Saharan Africa in recent years. Thus, we are left with an important analytical puzzle that requires explanation.

Such is the main contention in the article, but as part of an inventory of political violence in Africa it also reports several other findings. First, political violence in sub-Saharan Africa is not exceptional. While warfare and even mass killings have been central to Africa's post-colonial experience, in comparison to other world regions sub-Saharan Africa is not the leader in either the frequency or duration of such major forms of political violence. Second, consistent with the trends in warfare, large-scale mass killings of civilians, including genocide, are on the decline in Africa. Indeed, armed conflict and mass killing tend to correlate, thus the expectation is that as civil wars decline in size and intensity so too should the prevalence of mass killing and genocide decline. Third, other forms of political violence are salient in Africa, but they receive less analytical attention than warfare and mass killing. The article presents data on two forms of such violence – election-related violence and livelihood clashes, such as farmer–herder conflict, over access to land, water, and other local resources. The article contends that these forms of violence deserve greater visibility and theoretical attention than is generally accorded in the existing literature on political violence.

The article is a broad-gauged analysis. What it gives up in detail on specific cases it gains when looking at trends and patterns of political violence that are not often analysed together. Taken together, the findings present an image of contemporary Africa that challenges some common assumptions among non-specialists and even some specialists about the prevalence of violent conflict across the sub-continent. Rather than a place of endemic warfare, the trends suggest a sub-continent where large-scale organized political violence is on the decline and where large areas are stabilizing.13 War and other forms of large-scale violence are clearly not over, and the future may hold new outbreaks of major conflict. But the recent trends should be recognized and incorporated into empirical, theoretical, and policy accounts of Africa.

Trends in warfare

To measure longitudinal trends in warfare, as well as how Africa compares to other world regions, I turn to the leading dataset on civil wars – the Armed Conflict Database from the Uppsala Conflict Data Program (UCDP) and the Peace Research Institute of Oslo (PRIO).14 This database codes armed conflicts at a low threshold of 25 battle-related deaths per year in which there is the use of armed force between two parties, of which at least one is the government.15 The UCDP/PRIO data divide conflicts into three major types that are relevant to post-independent Africa: interstate armed conflict, internal state armed conflict, and internationalized internal state armed conflict.

Despite concerns about measurement error, coding decisions, and unit heterogeneity – concerns that are legitimate when coding large numbers of countries across multiple periods and relying on secondary sources16 – the broad-gauge analysis reveals patterns that are hard to detect otherwise. In particular, the UCDP/PRIO Armed Conflict data show that war, in particular civil war, has been a central feature of Africa's post-colonial political history. About thirty countries in sub-Saharan Africa – or around 65 percent of all states in the region – have experienced an armed conflict since independence.17 The basic trend over time is a steady accumulation of civil wars in the first three decades of independence, spiking in the early 1990s and then declining significantly (see Figure 1). The UCDP/PRIO data indicate that starting in the early 2000s there were on average eight to ten wars in any given year, which is about half the number of wars in sub-Saharan Africa in the early-to-mid 1990s.

Figure 1.

Armed conflicts in sub-Saharan Africa by type, 1960–2008.

Figure 1.

Armed conflicts in sub-Saharan Africa by type, 1960–2008.

There is considerable variation not only over time but also in the duration of these wars. There have been some very long wars, such as those in southern Sudan and in Angola, each of which lasted more than twenty years; there were also some short civil wars that lasted a few years, such as those in Mali in the early 1990s. There are places where wars are endemic, like Chad, Uganda, and Ethiopia with multiple wars of various durations since independence. But on average most wars are closer to the Tuareg/Arab rebellions in northern Mali than they are to the North–South civil war in Sudan. About 64 percent of African internal or internationalized internal armed conflicts in the UCDP/PRIO Armed Conflict dataset lasted five years or less; about 22 percent lasted 11 years or more. To be sure, those long wars are quite devastating socially and economically where they occur (as short wars can be, shown in Rwanda in the early 1990s), but the norm is not one of wars that never end.

In terms of frequency and duration, the UCDP/PRIO Armed Conflict data show that sub-Saharan Africa is not the most war-endemic region, and the region does not on average have the longest armed conflicts (see Figures 2 and 3). Effectively, those distinctions go to Asia, where – if we consider India, Afghanistan, Vietnam, the Philippines, Sri Lanka, and other countries – wars have been more prevalent and longer than in Africa, on average, over the 1960–2008 period. If one looks at the total number of wars divided by the total number of countries per region, the average in Asia is 1.88 wars per country, compared to 1.65 in sub-Saharan Africa and 1.14 in the Middle East.18 Africa has the largest number of armed conflicts, but also the largest number of countries, which makes the per-country averages for the region lower than in Asia.

Figure 2.

Mean duration of armed conflicts by type in five world regions, 1960–2008.

Figure 2.

Mean duration of armed conflicts by type in five world regions, 1960–2008.

Figure 3.

Armed conflicts per country by type in five world regions, 1960–2008.

Figure 3.

Armed conflicts per country by type in five world regions, 1960–2008.

A similar pattern holds with regard to duration. The average duration of internal, internationalized internal, and inter-state armed conflicts is longer in the Middle East and Asia than in sub-Saharan Africa. In fact, the UCDP/PRIO data show that Asian wars are almost twice as long on average as African wars. Divided by population, the results are different: Asia's population is overall about four times that of sub-Saharan Africa, so the per capita effects are greater in Africa. Nonetheless, contrary to common assumption, wars in Africa are neither longer nor more frequent on a per country basis.

Measured as a function of the intentional harm to civilians in wartime, the brutality of wars is harder to measure than frequency or duration. Yet some evidence suggests that African wars on average are not more brutal toward civilians compared to other regions. For example, Dara Kay Cohen's research on sexual violence finds that in civil wars since the early 1990s African conflicts have not featured more sexual violence than conflicts in other regions.19 The frequency and magnitude of mass killing in wartime was also less in Africa than in Asia from 1960 to 2008, as discussed below. That said, since the end of the Cold War, Africa has had greater proportions of such events as well as a greater proportion of cases of ‘one-sided killing’ than other regions.20

The purpose of this analysis is not to engage in crude regionalism. The main point is to acknowledge that armed conflict, especially civil war, has been central to Africa's post-colonial experience, and that Africa is and has been vulnerable to war and to atrocities. At the same time, the data run contrary to common assumptions, especially in non-academic settings but also in some academic circles, that Africa's proclivity to armed conflict, to long armed conflicts, and to atrocities in armed conflict, is unique and qualitatively distinct from other world regions.

Changing patterns of warfare

Quantitative measures of onset and duration paint an important picture of the overall trends in warfare at a general level, but they do not depict changing dynamics of war. Conventional wisdom is that civil wars in Africa have become more brutal in terms of violence against civilians, in terms of the presence of non-professional combatant forces, and in terms of the non-ideological goals of combatants in wars. This is the ‘new wars’ thesis.21 Reno's study of the evolution of warfare in Africa supports this general trajectory.22 If the Eritrean or South African freedom fighters were the faces of rebellion during the Cold War, warlords or génocidaires in Liberia, Sierra Leone, and Rwanda are the focal points of the past two decades. They represent armed thugs who fight to control diamond mines to get rich, who use child soldiers, who engage in unimaginable brutality against civilians, like cutting off civilians’ hands or committing mass murder. Gettleman's account claims that today's ‘un-wars’ in Africa are fought by atrocity-committing ‘predators’ who only want ‘cash, guns, and a license to rampage’.23 We are thus left with the empirical question of whether wars in Africa are more brutal, less ideological, and less professional than before?

I conclude there is a difference of degree but not of kind. Recent wars during the last two decades are almost always highly violent toward civilians, but the distinction should not be overstated. The real issue is that civil war is almost always brutal toward civilians, whether in Africa or elsewhere.24 Thus, wars of the past two decades in the DRC, Northern Uganda, Darfur, and Somalia, among others, are fought with enormous brutality and cruelty towards civilians, and combatants in these conflicts are generally not well-trained freedom fighters. But these dimensions of African civil wars are not entirely new. In the 1970s and 1980s, insurgent groups in Mozambique and Angola consistently used violence against civilians as a tactic. State-led mass killings took place in the same period – there was the Biafran civil war of the 1960s, Idi Amin and Milton Obote in Uganda in the 1970s and 1980s, the Burundi genocide of 1972, and the devastation unleashed upon the populations of southern Sudan by pro-government forces. Indeed, as shown below, the peak of large-scale mass killing episodes in sub-Saharan Africa was the 1980s, with a decline in the 1990s and 2000s. In short, when looking at atrocities in post-Cold War conflicts compared to atrocities in Cold War conflicts, there is a difference in degree, but not of kind.

The bigger change in the character of warfare is a decline of big wars that were fought for state control, that involved insurgents maintaining territorial control for long periods of time, and that pitted well-structured armies against each other. Writing in 2011, all the big wars in which two or more large, well-disciplined, well-armed, well-trained, hierarchical armies fought each other have ended. Wars of the early twenty-first century are typically smaller. They tend to have factionalized and divided armed insurgents; occur on the periphery of states; are difficult to end because of the mobile, factionalized armed groups; have strong cross-border dimensions; and to be wars where insurgents draw funding from illicit trade, banditry, and/or international terrorist networks rather than principally from major external states.

As examples, consider some of the conflicts that have occurred since 2005. These include the Ugandan state fighting the Lord's Resistance Army in northern Uganda; the Chadian and Central African Republic governments fighting insurgencies on their peripheries in the east; the Ethiopian government fighting in the Ogaden region; the Namibian government fighting against a separatist movement in the Caprivi Strip; the Sudanese government fighting rebel groups in Darfur (and Darfurian rebel groups fighting each other); Tuareg insurgencies in northern Mali and Niger; the Casamance insurgency in Senegal; the Cabinda separatists fighting in Angola; and Boko Haram in Nigeria, among other armed conflicts. These were small wars. The size of the insurgent armies is difficult to estimate accurately, but in most cases the insurgents resembled mobile armed bands whose numbers were comparatively few. Almost all these wars were fought on the peripheries of states. Many involved insurgencies with shifting and factionalized alliances – with Darfur, the DRC, Chad, and Senegal as good examples.

Of recent conflicts, the Democratic Republic of Congo might be considered an exception to this trend, given the involvement of seven state armies at the height of the conflict.25 But once those armies largely withdrew, eastern Congo resembled the broader trend: small bands of factionalized armed actors who control bits of territory, who engage in illicit trade, and who operate on the periphery, all the while causing suffering to the civilian population. So even the DRC, which at one point varied from the emerging norm, fits the pattern. Another possible exception was the brief war in Côte d'Ivoire that overthrew Laurent Gbagbo in April 2011. But even there the government's military forces largely stood down as the rebels advanced and, as noted earlier, the rebels were advancing in the context of broad domestic but especially regional and international opposition to Gbagbo. The rebels also ultimately collaborated with United Nations and French forces. In the end then, even Côte d'Ivoire did not experience a ‘big war’ – thus conforming to the trend highlighted here.

Evidence for the proposition is found in the declining level of battle deaths on the continent. In the most recent wars, actual combat fighting seems minimal even though civilian suffering frequently occurs. The UCDP/PRIO data show that, while battle deaths were never especially high in Africa, they have been on a steady decline.26 To derive the magnitude of battle deaths per conflict, Figure 4 shows the annual number of deaths divided by the number of civil wars in that year.27

Figure 4.

Battle deaths in civil wars in sub-Saharan Africa, 1960–2008.

Figure 4.

Battle deaths in civil wars in sub-Saharan Africa, 1960–2008.

In addition to the rise of smaller, factionalized insurgencies on the peripheries of states, there is evidence of a growing cross-border dimension to armed conflicts.28 To be sure, African wars in the Cold War had transnational dimensions. South Africa intervened directly in wars in its neighbouring states. Tanzania and Uganda fought each other in 1979, and Libyan forces invaded Chad in the same period. Neighbouring states often provided shelter and support to rebel organizations. Chadian and Angolan insurgents had rear bases in Sudan and the DRC, respectively. The Rwandan Patriotic Front invaded from Uganda. There are numerous other examples from the Cold War period and just afterwards.

Yet a largely new development is mobile insurgent groups that move back and forth across national borders, engage the security forces of multiple states, and inflict harm on civilian populations across borders. The Lord's Resistance Army typifies the pattern; in recent years, the LRA has moved bands of fighters across northern Uganda, Sudan, the Democratic Republic of Congo, and the Central African Republic. Al-Qaeda in the Maghreb (AQIM) is another example. AQIM moves across the Sahel – in Algeria, Mali, Niger, and Mauritania. Even Al-Shabab, which is based in Somalia, has launched operations in Uganda and Kenya – for example, the 2010 World Cup bombing in Kampala – and both Ethiopia and Kenya have intervened recently in Somalia to contain the Islamists. The first and second wars in the DRC, which were clearly bigger wars than those just described, nonetheless followed only a slightly different model; in those cases, rebel groups from Rwanda, Burundi, Angola, and Uganda had rear bases in the DRC. What differed, compared to other regions, is that those neighbouring states in turn intervened and remained in the DRC for significant periods of time.

Taken together, these varied findings both challenge and support some prevailing views on African warfare. The findings strongly suggest a significant change in the character of warfare in Africa, namely a shift from large wars to smaller ones on the peripheries of states. The findings also run contrary to seeing civil wars as nationally contained conflicts; they increasingly have trans-border dimensions. The findings also support some aspects of the ‘new wars’ thesis, holding that post-Cold War conflicts involve factionalized and unprofessional insurgencies as well as atrocity-committing combatants more likely to predate on civilians than to fight each other. At the same time, the distinction should not be overdrawn –the so-called ‘old’ wars of the Cold War (and before) involved atrocity-committing insurgencies and states.

Trends in mass killing and genocide

Although war, on the one hand, and genocide and mass killing, on the other, are related – the latter most frequently occurs when the former is present29 – the frequency of the two forms of political violence in Africa are generally analysed separately. In this section, I look at trends in large-scale intentional killing of civilians over time to see whether or not they track changing patterns of warfare on the continent. Even though detailed disaggregated data on intentional mass killing of civilians in Africa are weak, a US government-funded agency, the Political Instability Task Force (PITF), has initiated a data collection project on mass killing events during the past 60 years.30 The PITF data measure events in which at least 1,000 intentional deaths have occurred during a continuous period. Because of concerns about data quality and because I want to focus on mass killing events at a higher magnitude, I restrict cases to those in which at least 10,000 civilians were deliberately killed, according to the PITF data.

As with warfare, compared to other world regions the data show that sub-Saharan Africa is not the region most prone to mass killing. From 1960 forward, there are some 20 distinct episodes in sub-Saharan Africa (with a total of 46 states), compared to 22 episodes in Asia (with a total of some 34 states), and eight episodes in the Middle East (with a total of 22 states). Thus, neither in relative nor in absolute terms is sub-Saharan Africa the regional leader in terms of incidents of mass killing in which 10,000 civilians were killed.

Looked at longitudinally, there is consistency in the frequency of episodes of mass killing per decade, with a peak in the 1980s. During the 1960s, there are six episodes recorded in the data: Sudan (first war), Rwanda, Ethiopia (in Eritrea), Burundi (1965), Nigeria (Biafra), and Equatorial Guinea. In the 1970s, there were eight, including some episodes that started in the 1960s: Sudan (first war), Ethiopia (in Eritrea), Ethiopia (Dergue political killings), Ethiopia (in Ogaden), Burundi (1972), Equatorial Guinea, Uganda (Idi Amin), and Angola. In the 1980s, there were nine episodes: Uganda (Obote II), Zimbabwe, Somalia, Chad, Sudan (second war), Ethiopia (in Eritrea), Burundi (1988), Ethiopia (in Ogaden), and Angola. In the 1990s, there were six: Ethiopia (in Eritrea), Sudan (second war), Burundi, Rwanda, Democratic Republic of Congo, and Angola. And in the 2000s, there were five: DRC, Sudan (second war), Darfur, Burundi, and Angola.31

In each case, these incidents took a terrible toll and caused enormous civilian suffering. Yet, consistent with the trend in civil wars, there is a significant decline in recent years from the peak in the 1980s. Indeed, if one were to extend the data outwards to the start of 2010, three of the five cases were no longer active mass killing episodes; in other words, by the start of 2011, only Darfur and the DRC remained cases with active mass civilian destruction. As with large wars, mass killing and genocide are on the decline in Africa.

Other forms of political violence

The literature on violence is dominated by studies of war and mass killing. Yet other forms of violence in Africa (and beyond) deserve greater study. The Uppsala Conflict Data Program has developed one dataset, of ‘non-state armed conflicts,’ which measures conflict between two armed groups that are not the state and that result in 25 deaths.32 Using these data, Williams counts 287 dyads in Africa between 1990 and 2009, most of which occurred in the DRC, Ethiopia, Kenya, Nigeria, Somalia, Sudan, and Uganda.33 While important and valuable, however, the dataset groups together a number of different contexts of conflict. Some are clearly connected to civil war, as in intra-rebel, paramilitary–rebel, or inter-clan fighting in Somalia, Senegal, the DRC, and Sudan. Other dyads measure fighting between workers’ associations, as in South Africa. Still others are related to inter-communal violence (Nigeria), to elections (Kenya), and control of resources (Ethiopia).34

Building on the insight that other forms of violence matter besides state-based armed conflict and mass killings, but choosing a different way to categorize them, I argue that two forms of violence are especially salient in contemporary sub-Saharan Africa. The first is electoral violence – that is, violence directly associated with an electoral contest, either before, during, or after a poll has taken place – and localized conflict over access to livelihood resources, in particular land and water. While the topics have received some attention, particularly with regard to specific cases such as Kenya and Zimbabwe on electoral violence35 and Mali and Côte d'Ivoire on the resource conflicts,36 these forms of violence are to date significantly understudied compared to civil war and mass killing/genocide.

Preliminary evidence indicates that both forms of violence matter a great deal in sub-Saharan Africa and even more so in recent years. This is clearly the case for electoral violence. During the 1970s and 1980s, electoral violence was rare to non-existent because one-party regimes were dominant (though in the 1960s after independence some elections triggered violence). Yet the reintroduction of multi-party contests in the 1990s saw an increase of such violence. Many Africanists know electoral contests can be violent, given recent high-profile examples in Zimbabwe, Kenya, and Côte d'Ivoire. Yet, in contrast to the research on civil wars and mass killing, there has been little systematic cross-national data collection on the frequency and variation of levels of electoral violence in Africa.

To address that gap in the literature, Charlie Taylor and I created the African Electoral Violence Database (AEVD).37 The database measures the incidence of electoral violence, the level of electoral violence, whether the electoral violence occurred before or after the election, and whether incumbents or challengers [or both] were the main actors. Among the pertinent central findings from the data analysis are that high levels of electoral violence are comparatively rare (about 20 percent of cases, N = 221 elections) and that the trend line since the end of the Cold War has stayed relatively constant, as Figure 5 shows.38 The AEVD assigns a 0–3 score, with zero being no violence, one being intimidation and harassment, two being targeted assassinations, and three being 10 or more civilian deaths.

Figure 5.

Electoral violence in sub-Saharan Africa, 1990–2008.

Figure 5.

Electoral violence in sub-Saharan Africa, 1990–2008.

As Figure 5 shows, the mean level of violence in African elections was relatively constant between 1990 and 2008. The data show that about 42 percent of elections had no reported electoral violence (category 0), 38 percent had violent harassment (category 2), and 10 percent in each case produced targeted assassinations or more than 10 deaths (categories 3 and 4). Thus, while about 60 percent of African elections in the dataset see violent intimidation or worse, only about 20 percent of the elections involve high levels of violence. Moreover, given that the trend line has stayed constant, and multi-party contests remain the norm, the expectation is that this form of political violence is likely to persist – at least in the short term. More research into electoral violence in Africa is needed.

The second major form of political violence that is salient and likely to persist concerns localized conflicts over access to critical livelihood resources, such as farmer–herder conflicts. Cross-national and longitudinal data on the frequency and duration of this form of violence are limited, in part because event-based datasets such as the UCPD dataset record incidents when a threshold of twenty-five or more are killed. By contrast, often these conflicts result in low-level fatalities, such as a death or two, though they can escalate to much more violent wars. Yet country-specific evidence indicates that lethal and disruptive violence over access to resources can be persistent and acute.39

Country-specific evidence is supported by cross-national data collection as well. A newly available dataset called ‘Social Conflict in Africa’ records incidents of violent conflict covered in international news accounts (see Figure 6). The dataset records whether the news account mentions ‘food, water, subsistence’ or ‘environmental degradation’ as a source of the conflict. While the quality and depth of international news reporting on rural Africa is notoriously weak, the data collected provide evidence of growing numbers of conflicts over livelihood resources.40

Figure 6.

Violence over access to livelihood resources in sub-Saharan Africa, 1990–2009.

Figure 6.

Violence over access to livelihood resources in sub-Saharan Africa, 1990–2009.

It is difficult to know if the greater recent frequency of these conflicts is a product of new interest in the phenomenon, which in turn would lead to more attention and reporting. Nonetheless, the main point is that country-based evidence and cross-national evidence suggest that violent conflicts over access to vital resources are critical forms of political violence that deserve greater empirical and theoretical attention.

A key question is whether/how these forms of violence causally interact. In the electoral violence cases of Kenya, Zimbabwe, and Côte d'Ivoire, for example, access to land has been a central dimension to the violence that occurred. In civil war cases, such as the Democratic Republic of Congo, competition over access to local resources has given rise to claims of ‘autochthony’ in the context of warfare.41 David Laitin and James Fearon report that for nearly a third of all ‘ethnic’ civil wars in their cross-national dataset on civil war the spark was a ‘sons of the soil’ conflict. These are where a regional group that considers itself indigenous clashed with a migrant group.42 These observations suggest that different forms of political violence are interlinked, as war and mass killing are. These connections also deserve greater attention.

Puzzles and theory

What explains the decline of civil wars and the shift from larger civil wars to smaller, more peripheral civil wars in Africa? The change is a puzzle. Most leading theories explaining the onset of civil war privilege domestic variables that are slow-moving or invariant across time, such as weak states, rough terrain, oil and high-value mineral endowments, and horizontal inequalities.43 These causal variables are persuasive; there is strong face validity to the claim that they are individually or jointly responsible for the onset of many civil wars in Africa. Yet these factors cannot sufficiently account for the sharp decline of warfare in Africa and its changing character.

Domestic factors such as stronger civil societies, consistent economic growth, or stronger states are good candidates to explain the change. But following some recent work on global patterns of warfare,44 I emphasize geo-political shifts. These include a decline in external state support for insurgencies, the promotion of multi-party elections, significant investments in conflict prevention and mediation after the Cold War, and the rise of China.

During the Cold War, the United States and the Soviet Union were major sources of funding for insurgencies and states fighting insurgencies. From the Horn of Africa to the southern states, the superpower rivalry meant that states and insurgencies had access to weaponry, training, ideological discipline, and diplomatic support. When the Cold War ended, some states that had received previous external support during the Cold War became newly vulnerable, such as Mengistu's Ethiopia, Siad Barre's Somalia, Mobutu Sese Seko's Zaire, and Liberia's Samuel Doe. By the same token, some rebel groups that had received funding through Cold War channels became weaker, such as UNITA in Angola. In other locations, the end of the Cold War created a new window of opportunity for armed opposition groups who had been waiting for the right moment to start their insurgencies, as in Rwanda and Mali. Thus, the first decade following the end of the Cold War saw an immediate increase in warfare, as belligerents saw new conditions and opportunities to start or settle conflicts.

But thereafter the frequency of wars declined and their character changed. Beginning in the late 1990s, the external opportunities for insurgents to garner weaponry, training, advisory input, and ideological discipline became much more meagre. States such as Sudan supported insurgencies to disrupt their neighbours, as with the Lord's Resistance Army, but large-scale international support for insurgencies as structured fighting forces and governments in waiting sharply declined. New insurgencies started across the continent, and in places they survived because of access to mineral resources or because the states they fought were weak. But these insurgencies did not develop into the kinds of well-structured guerrilla armies that evolved during and just after the Cold War.

A related change is the rise of multi-party electoral rules that followed the end of the Cold War. On balance, the opening of the electoral terrain, however flawed in some cases, attracted would-be insurgents away from the lure of the bush and toward the political arena. The onset of multi-party elections meant that, from a would-be insurgent's point of view, governments were at least nominally vulnerable outside the context of armed resistance. Moreover, the weight of international funding flowed toward sponsoring elections and civil society organizations. For talented opposition figures, the opening of the political arena – combined with the change in international funding streams – created a strong pull away from the battlefield toward the domestic political arena.

Many insurgencies that persisted in the last decade are counter-system rebellions. The rebel leaders seek to change the fundamental rules of the game, as is the case with the AQIM, Al Shabab, the LRA, and separatists in Namibia and Senegal. In other words, those insurgencies that survive cannot press their case electorally because they seek fundamentally different political set-ups, which is indirect evidence that democratization has in fact sapped some of the would-be insurgents away from the bush. Even in Côte d'Ivoire, the insurgency that began in 2002 and continued in 2010 was directly related to manipulations of the electoral playing field. When a segment of the population was categorically excluded from electoral contests in the early 2000s and when the incumbent sought to cancel results in 2010, insurgents fought. But on balance the Ivoirian rebels were not a fighting force designed to overthrow the Gbagbo regime but rather to force inclusive elections.

The geo-political shifts after the Cold War saw two additional, related changes. One is the increasing presence of China on the continent. In Africa and elsewhere, China's foreign relations philosophy is non-interference in the domestic affairs of states. In practice, while China has provided support to non-democratic states in exchange for access to resources, China has not, as far as current evidence suggests, favoured funding insurgencies.45 This too represents another shift from the Cold War, one that narrows the opportunities for rebel support, diminishes the likelihood that small insurgencies can grow into large-scale fighting forces, and creates incentives for talented opposition figures to remain in the domestic political space.

Lastly, the end of the Cold War has seen a strengthening of international and regional mechanisms of dispute resolution and conflict containment.46 In the African context, three mechanisms are salient. The first is United Nations peacekeeping. As Figure 7 shows, UN peacekeeping has grown remarkably in the size of missions in sub-Saharan Africa during the last twenty years and in the sophistication of mandates.47 In the past decade, almost every UN peacekeeping mission is deployed with a robust Chapter VII mandate, as opposed to more limited Chapter VI mandates, which were dominant in the 1990s and before.

Figure 7.

United Nations peacekeeping operations in sub-Saharan Africa, 1990–2010.

Figure 7.

United Nations peacekeeping operations in sub-Saharan Africa, 1990–2010.

To be sure, one should not be naïve about the problems of peacekeeping.48 On the other hand, there has been learning in the organization, and empirically the missions are more frequent, more sophisticated, and larger than at any previous time. There are sound theoretical reasons to think UN peacekeeping may have an effect on conflict reduction.49

Second, African regional mechanisms are stronger. The African Union and regional organizations like ECOWAS have taken a greater interest in the prevention of armed conflict.50 In addition, in many cases African luminaries serve as ad hoc ambassadors of peace. In Kenya's 2007–8 crisis, for example, Kofi Annan was a key player; in Côte d'Ivoire's 2010–11 crisis, Raila Odinga and five African heads of state were quite active. South Africa plays an increasingly active role in settling African wars. Jerry Rawlings has emerged as an elder statesman. Critics question the decisions and effectiveness of these actors, but on balance their presence is greater than in previous periods of African international relations.

Finally, international criminal justice mechanisms – from the ad hoc tribunals for Rwanda and Sierra Leone to the International Criminal Court – are stronger than ever before.51 Again, one may be circumspect about the strong deterrence claims of advocates, but on the other hand in some cases the implantation of these mechanisms correspond with a decline of warfare.

On balance, it is reasonable to conclude that each of these mechanisms shapes the incentives – even marginally – of African decision makers in ways that limit the extent of warfare. They may not always work in every case, but they point to a stronger and strengthening international conflict reduction regime that has emerged since the end of the Cold War. The proposition clearly deserves greater empirical testing. But taken together with the diminished opportunities and incentives for civil war that followed the end of the Cold War, we have evidence to explain the recent change in frequency and character of warfare observed during the past decade.

Conclusions

I have argued that political violence has been fundamental to Africa's post-colonial political history. But I have also argued that Africa is not uniquely prone to violence, and in fact – given Africa's colonial history, weak states, ethnic difference, and economic endowments – we might have expected more civil war (and mass killing) than has been the case. I have also argued that war is not over, but the most recent trend is a sharp decline in frequency and a change in character. At the time of writing, there are almost no big wars left on the continent. Those wars that remain most commonly entail factionalized insurgents operating on the peripheries of states, and these insurgencies rarely hold substantial territory, involve small numbers of fighters, and exhibit strong transnational connections. Given these characteristics, one should expect these smaller wars with factionalized, marginalized insurgencies to persist – the insurgencies will be hard to contain, and their factionalized nature creates difficulty in terms of negotiated settlements. Many groups also maintain what I have called ‘counter-system’ demands. Finally, I have claimed that new forms of political violence that hitherto have received comparatively less attention are likely to persist in coming years – namely, electoral violence and violence over access to critical livelihood resources. Both deserve greater empirical and theoretical study as do other forms of political violence.

The macro-level changes in patterns of political violence, especially the sharp decline of major wars, constitute an important analytical puzzle that is not well explained in the existing literature. While the article is primarily descriptive and empirical, I have argued that external, geo-political changes account for the changing frequency and character of warfare. In particular, I argue that the end of the Cold War dried up sources of funding for insurgencies and states to fight proxy wars: while in the short term (in the 1990s), this created windows of opportunities for old conflicts to be settled or new conflicts to start, the change over time meant that the conditions that favoured the creation of big wars involving well-supplied and structured armies declined. The end of the Cold War also created new opportunities in the electoral arena, and thus new incentives for opposition leaders to contest for power through non-military means, diverting their ambitions and resources from bush-based strategies. Two other, related changes reinforced these dynamics. The first is the growing presence of China on the continent, which at least to date is not apparently providing support to insurgencies. The second is a strengthening of international and regional mechanisms of conflict prevention and mediation. All of these hypotheses deserve greater testing than the narrative account provided here.

The article does not address in detail why large-scale mass killing and genocide is on the decline or why electoral violence and violence over access to livelihood resources are persistent (or increasing). Given the strong empirical correlation between war and mass killing, the decline in wars is probably responsible for the decline in mass killing events. But why electoral violence and why violence over access to livelihood resources persist – and what explains variation in these forms of violence across African states – are not addressed here. These questions clearly deserve greater research, given the likely continued salience of these forms of violence.

A final caveat must be made: writing about the past does not predict the future. Major wars involving large well-organized armies could break out in Africa, and mass killing and genocide events could increase again in frequency. Several areas of the continent remain volatile, including but not limited to eastern DRC, Sudan, Somalia, and the Sahelian space in northern Mali and Niger.52 But if the most recent trends continue and if recent geo-political changes are responsible for the recent trends, we should expect a continued shift away from large-scale armed conflict to smaller, more peripheral wars, a continued decline of mass killing, and a continuation of other forms of less lethal but important political violence, such as electoral violence and local violence related to access to vital resources. That suggests a picture of a post-Cold War Africa where violence is still real, but an image sharply at odds with that of endemic, ‘forever’ warfare and an empirical reality that deserves widespread recognition among scholars and policy makers who concentrate on Africa.

1
Throughout the text and in the graphs I refer to ‘sub-Saharan Africa’ and ‘Africa’ interchangeably. Those states under discussion in the article are those south of the Sahara.
2
Robert Kaplan, The Coming Anarchy: Shattering the dreams of the post-Cold War (Random House, New York, NY, 2000). Note that the title page displays a map of Africa.
3
Jeffrey Gettleman, ‘Africa's endless wars: why the continent's wars never end’, Foreign Policy 178 (March/April 2010), pp. 73–5.
4
Paul Williams, War and Conflict in Africa (Polity Press, Cambridge, 2011), p. 4. Williams explains the African focus of his insightful and perspicacious book, citing ‘the sheer volume of armed conflicts’ on the continent since the Cold War.
5
By civil wars I refer to armed conflict within an internationally recognized country between two or more belligerents in which one actor is the state.
6
Joshua Goldstein, Winning the War on War: The decline of armed conflict worldwide (Penguin Books, New York, NY, 2011); Human Security Report Project, Human Security Report 2009/2010: The causes of peace and the shrinking costs of war (Simon Fraser University, Vancouver, 2010); John Mueller, The Remnants of War (Cornell University Press, Ithaca, NY, 2004).
7
Williams, War and Conflict, p. 4.
8
Thomas Bassett, ‘Winning coalition, sore loser: the 2010 presidential elections in Côte d'Ivoire’, African Affairs110, 440 (2011), pp. 457–68.
9
Again contra Gettleman who claims today's rebels have no ideology and no cause, many of these insurgents make strong ideological claims, such as Boko Haram, AQIM, Al-Shabab, Darfuri rebels, and the various separatist movements.
10
William Reno, Warfare in Independent Africa (Cambridge University Press, New York, NY, 2011).
11
To be sure, some insurgencies fight in large, weak, personalized states, as in Chad, the Central African Republic, and the Democratic Republic of Congo.
12
These theories are discussed in more detail below.
13
The latter point is consistent with a major forthcoming study: Crawford Young, The Post-Colonial State in Africa: Fifty years of independence (University of Wisconsin Press, Madison, WI, forthcoming).
14
Nils Petter Gleditsch, Peter Wallensteen, Mikael Eriksson, Margareta Sollenberg, and Håvard Strand, ‘Armed conflict 1946–2001: a new dataset’, Journal of Peace Research39, 5 (2002), pp. 615–37. The data are available here: <http://www.prio.no/CSCW/Datasets/Armed-Conflict/UCDP-PRIO/> (25 January 2012). The dataset used in the analysis is v4_2009.
15
In the discussion and analysis, I treat any ‘armed conflict’ in the UCDP/PRIO Armed Conflict database as a war. (There is also a separate database on non-state armed conflict, which is discussed below.) I choose the lower threshold of 25 battle deaths per year (as opposed to 1,000 battle deaths per year) so as to be sensitive to the occurrence of both high- and low-intensity wars.
16
Nicholas Sambanis, ‘Using case studies to expand economic models of civil war’, Perspectives on Politics2, 2 (2004), pp. 259–79. Paul Williams in War and Conflict (Chapter 1 in particular) persuasively shows the ways in which coding decisions and measurement error lead to inconsistent results. Nonetheless, despite these very real and important concerns, by using a well-respected and well-researched dataset such as the UCDP/PRIO one can provide a valid snapshot of trends over time.
17
These data exclude anti-colonial wars, which are coded as ‘extra-systemic’ wars in the UCDP/PRIO dataset. The baseline number of sub-Saharan African states in the quantitative analysis is 46, which excludes the Seychelles and São Tomé and Principe.
18
Several African states (such as Zimbabwe, Angola, Mozambique, Namibia, and Eritrea) achieved their independence after the early 1960s, whereas few states in Asia achieved independence in the comparable period (Bangladesh and East Timor). Thus, the number of years in which some African states could have had a war was less overall than in Asia, which is worth noting.
19
Dara Kay Cohen, Explaining Sexual Violence during Civil War (Stanford University, unpublished PhD thesis, 2010).
20
Williams, War and Conflict, pp. 4, 26. One-sided killing connotes the idea that unarmed civilians or prisoners are targeted.
21
Mary Kaldor, New and Old Wars: Organized violence in a global era (Stanford University Press, Palo Alto, CA, 1999). For a challenge to that idea, see Stathis Kalyvas, ‘“New” and “old” wars: a valid distinction?’, World Politics54, 1 (2001), pp. 99–118.
22
Reno, Warfare in Independent Africa.
23
Gettleman, ‘Africa's forever wars’.
24
Stathis Kalyvas, The Logic of Violence in Civil War (Cambridge University Press, New York, NY, 2006).
25
Filip Reyntjens, The Great African War: Congo and regional geopolitics, 1996–2006 (Cambridge University Press, New York, NY, 2009).
26
The UCDP/PRIO data record high, low, and best estimates of battle deaths per war per year. Figure 4 reports the high estimates in the data. The data are available at <http://www.prio.no/CSCW/Datasets/Armed-Conflict/Battle-Deaths/> (25 January 2012). The data used in this article are version 3.0. See Bethany Lacina and Nils Petter Gleditsch, ‘Monitoring trends in global combat: a new dataset of battle deaths’, European Journal of Population21, 2–3 (2005), pp. 145–66.
27
The data are on civil war, which is the dominant form of armed conflict on the continent. Yet the trends for interstate and internationalized civil war are similar.
28
Idean Salehyan, Rebels without Borders: Transnational insurgencies in world politics (Cornell University Press, Ithaca, NY, 2009).
29
Scott Straus, ‘Second-generation comparative research on genocide’, World Politics59, 3 (2007), pp. 476–501; Jay Ulfelder and Benjamin Valentino, ‘Assessing the risks of state-sponsored mass killing’ (Political Instability Task Force, Washington, DC, 2008).
30
Ulfelder and Valentino, ‘Assessing the risks’.
31
The data cover 2000–8.
33
Williams, War and Conflict, p. 23.
34
Williams produces a very readable appendix of the conflicts in Africa; see War and Conflict, pp. 235–42.
35
Catherine Boone, ‘Electoral populism where property rights are weak: land politics in contemporary sub-Saharan Africa’, Comparative Politics41, 2 (2009), pp. 183–201; Susanne Mueller, ‘The political economy of Kenya's crisis’, Journal of Eastern African Studies2, 2 (2008), pp. 185–210.
36
Thomas Bassett, ‘The political ecology of peasant–herder conflicts in northern Ivory Coast’, Annals of the Association of American Geographers78, 3 (1988), pp. 453–72; Matthew Turner, ‘Political ecology and the moral dimensions of “resource conflicts”: the case of farmer–herder conflicts in the Sahel’, Political Geography23, 7 (2004), pp. 863–89.
37
Charlie Taylor is an advanced graduate student in the Department of Political Science at the University of Wisconsin, Madison. Leonardo Arriola, at the University of California, Berkeley, is also working on a dataset of electoral violence in Africa.
38
Scott Straus and Charlie Taylor, ‘Democratization and electoral violence in sub-Saharan Africa, 1990–2008’ in Dorina Bekoe (ed.), Electoral Violence in Sub-Saharan Africa (United States Institute of Peace, Washington, DC, forthcoming).
39
For a good review of the literature, see Mark Moritz, ‘Understanding farmer–herder conflicts in west Africa: outline of a processual approach’, Human Organization69, 2 (2010), pp. 138–48.
40
The data are here: <http://ccaps.strausscenter.org/scad/conflicts> (25 January 2012).
41
Morten Boas, ‘“New” nationalism and autochthony: tales of origin as political cleavage’, Africa Spectrum44, 1 (2009), pp. 19–38.
42
James Fearon and David Laitin, ‘Sons of the soil, migrants, and civil war’, World Development39, 2 (2011), pp. 199–211. Four African civil wars in their dataset are ‘sons of the soil’ conflicts – Mali (Tuareg rebellion), Senegal (Casamance), and Sudan (Darfur). They also include the anti-colonial war in Zimbabwe, though anti-colonial wars are not the focus of the discussion in this article.
43
On weak states and rough terrain, see James Fearon and David Laitin, ‘Ethnicity, insurgency, and civil war’, American Political Science Review97, 1 (2003), pp. 75–90; on economic endowments, see Paul Collier and Anke Hoeffler, ‘Greed and grievance in civil war’, Oxford Economic Papers56, 4 (2004), pp. 563–95; on horizontal inequalities, see Lars-Erik Cederman, Nils Weidmann, and Kristian Skrede Gleditsch, ‘Horizontal inequalities and ethnonationalist civil war: a global comparison’, American Political Science Review105, 3 (2011), pp. 478–95 and Frances Stewart (ed.), Horizontal Inequalities and Conflict: Understanding group violence in multiethnic societies (Palgrave MacMillan, New York, NY, 2008).
44
Stathis Kalyvas and Laia Balcells, ‘International system and technologies of rebellion: how the end of the Cold War shaped internal conflict’, American Political Science Review104, 3 (2010), pp. 415–29.
45
On China's foreign policy in Africa as it relates to coups and political violence, see Jonathan Holslag, ‘China and the coups: coping with political instability in Africa’, African Affairs110, 40 (2010), pp. 367–86.
46
This is a point emphasized in Goldstein, Winning the War and Human Security, Human Security Report. See also Jon Western and Joshua Goldstein, ‘Humanitarian intervention comes of age’, Foreign Affairs90, 6 (2011), pp. 60–73.
47
The data are from official UN peacekeeping statistics, found at <http://www.un.org/en/peacekeeping/operations/past.shtml> and <http://www.un.org/en/peacekeeping/operations/current.shtml> (25 January 2012). The data include military personnel but exclude police and civilian staff.
48
Séverine Autesserre, The Trouble with the Congo: Local violence and the failure of international peacebuilding (Cambridge University Press, New York, NY, 2010); Michael Barnett and Martha Finnemore, Rules for the World: International organizations in global politics (Cornell University Press, Ithaca, NY, 2004).
49
Virginia Page Fortna, Does Peacekeeping Work? Shaping belligerents' choices after civil war (Princeton University Press, Princeton, NJ, 2008); Lise Howard, UN Peacekeeping in Civil Wars (Cambridge University Press, New York, NY, 2009).
50
Alex Bellamy and Paul Williams, ‘The new politics of protection? Côte d'Ivoire, Libya, and the responsibility to protect’, International Affairs87, 4 (2011), pp. 845–70; Thomas Bassett and Scott Straus, ‘Defending democracy in Côte d'Ivoire: Africa takes a stand’, Foreign Affairs90, 4 (2011), pp. 130–40.
51
Victor Peskin, International Justice in Rwanda and the Balkans: Virtual trials and state cooperation (Cambridge University Press, New York, NY, 2008).
52
Indeed, as the article went to press, a new and significant Tuareg offensive began in northern Mali. Equipped with heavy weapons from Muammar Gaddafi's fallen regime in Libya and commanded by a former colonel from Libya's army, the insurgents successfully fought against Malian troops and captured territory in the north during the first weeks of fighting. What will happen remains unclear, and the case may diverge from the trends described in the article, even if to date the fighting has been confined to the far north. The timing and initial success of the insurgents is consistent with the theoretical claim in the article that external, non-domestic changes (in this case regional) affect the dynamics of internal armed conflict.

Comments

0 Comments