Draft: Please do not quote or cite

African Militaries and Rebellion:

The Political Economy of Threat and Combat Effectiveness

Jeffrey Herbst*

Paper to be presented at the conference on “Civil War Duration and Post-Conflict Transitions,” University of California, Irvine, 18-20 May 2001.

Professor of Politics and International Affairs, Bendheim Hall, Princeton University, Princeton, NJ 08544. Phone: 609 258-2160. Fax: 609 258-4772. Herbst@Princeton.Edu

1

War in Africa, and in other parts of the world, has recently attracted

significant attention given the cost of those conflicts in human lives and damage

to national and regional economies. A growing and p roductive literature has

emerged focussing on the motivations of rebels during these wars. Important

econometric work has attempted to explain the economic and political

motivations of rebels (Collier and Hoeffler: 1998); there have been case studies

of different guerilla movements (e.g., Clapham: 1998); and cross-national

analysis of the organization of rebellion (Herbst: 2000a). However, there has

been no corresponding literature on, quite literally, the other side: the political

economy of how and why national militaries perform during civil war. This paper

will examine the political and economic determinants of how African militaries

face the threat of rebellion and of differential levels of effectiveness in combating

insurgents. It will thus attempt to complement the literature that has recently

emerged on rebellion with what might be called an examination of the industrial

organization of counter-insurgency.

The paper argues that the particular threat posed by insurgency in Africa

must be understood in order to analyze how militaries operate. By examining the

life cycle of rebellion in Africa, I conclude that some economic models of the

relationship between investments in national armies and combat cabilities are not

applicable to the way wars are actually fought. In particular, it is difficult to

associate investments in the military with force projection because it may only be

possible early in an insurgency to defeat rebels through military means.

However, few African armies are able to mobilize quickly enough to defeat an

2

insurgency in its early phases. At the same time, who the enemy is—especially

whether citizens are asked to fight other citizens or foreigners—has an impact on

the ability to mobilize in the face of a threat. Yet the characteristics of the enemy

are often ignored when trying to model the military.

The Underdeveloped Study of Militaries

The failure to study militaries as important actors in rebellions is

surprising. Indeed, the most obvious question concerning almost all the civil

wars in Africa is: Why have militaries performed so poorly given that the actual

armed threat post by rebels (who are often poorly organized, employ large

number of child soldiers, and are inadequately funded) can best be described as

pathetic? For instance, the Revolutionary United Front in Sierra Leone, while

posing a persistent threat to the government in Freetown and adept at brutalizing

civilians, was never much of a fighting force and was effectively deterred by the

appearance of a few hundred British Marines and special forces units. Yet, very

few African armies have won outright victories against rebels or have been able

to change the military facts on the ground so that rebels would have to sue for

peace.

Indeed, the conduct of militaries during civil wars is a critical determinant

of the duration and course of conflict. Young makes the obvious point: “military

factors, although often neglected by academics, have often determined the

ultimate outcome of modern African insurgencies (1996, p. 179).” Similarly,

Russell noted in one of the rare systematic studies of rebellion “No mass

rebellion can succeed without the defection of some of the regime’s armed

3

forces. . . In a situation where people are rebelling, the behavior of the armed

forces has been shown to be a decisive factor in the outcome of the rebellion.

For revolutionaries to come to terms with this means that they must devote a

great deal of thought to how to encourage defections from the police and the

army (1974, p. 87).” Similarly, Herbst (2000a) finds that the degree of threat post

by the armed forces of the state is a critical determinant of how rebels will

structure their own forces.

Unfortunately, beyond specialists writing for other security analysts, there

is only a poorly developed literature on African militaries qua militaries. Most of

the analysis on African militaries has focused on their roles in coups or as

leaders of government because most armies on the continent have interfered far

more often with their own political system than have fought on the battlefield.

Indeed, relatively few African militaries have had any experience in combat

although that observation is less true than it was a generation ago given the

number of African countries that are now at war. Thus, even the most critical

questions regarding the competence of national militaries to fight have only been

examined in a cursory manner. For instance, as William Thom notes, “The ability

of African armies to expand their force levels in response to hostilities is not well

understood, mostly because little is known about military reserves and

mobilization systems. The response is to rate mobilization as very low if not nil

(Thom: 1988, p. 20).” More generally, Snyder claims that “There is no well-

defined relationship between a nation’s security and its own and other states’

military expenditures (Snyder: 1986, p. 119).

4

Understanding African Militaries

There are hints in the literature that militaries do not only fight for the

public good but also have economic agendas (Reno: 2000). However, there has

been no systematic study of how militaries operate from a political economy

perspective. The implicit assumption in the literature has been that while rebels

may have a host of motivations beyond their stated political and ideological aims,

militaries in Africa and elsewhere defend the state. For instance, Grossman

argues, “The function of the soldiers is to protect the income of the ruler’s

clientele either by deterring the potential insurgent leader from organizing an

insurrection or by counteracting actual insurgents (Grossman: 1995, p. 194).”

Olson has in many ways provided the best description of why rulers

provide security. In his model, if a leader is strong enough to hold territory and

monopolize theft--what Olson calls a “stationary bandit--” he has an interest in his

domain prospering. Olson argues that, “This encompassing interest leads him to

[the leader] . . . spend some of the resources that he controls on public goods

that benefit his victims no less than himself. . . he prohibits the murder or

maiming of his subjects. . . He serves his interests by spending some of the

resources that he controls to deter crime among his subjects and to provide other

public goods (Olson: 2000, p . 10).” This is an important explanation of how a

secure environment develops without having to assume good intentions on the

part of the leader. However, Olson does not suggest how even a leader with

significant interests in providing security can actually go about doing it. Indeed,

Olson appears to make a series of unexplored assumptions about how leaders

5

can keep control over their security forces and have them continue to fight for

him when there are clear private gains to be made by the men with guns if they

follow their own agenda and not the leaders.

It is generally assumed that armies can be modeled using an easily

recognizable function that relates competence in combat to the inputs allocated

to the military. Thus, in Hirshleifer’s model, there is a “combat power function” in

which resources devoted to combat readiness determine success in

appropriation (1988, p. 207). Similarly, Polo describes a “black box” where, “We

assume military capacity to be an increasing function of the number of members

of the organization: all the relevant functions—monitoring, intelligence and

violence—require labor as a fundamental input (Polo: 1995, p. 89).” However, it

is important to note that Polo also includes a spatial dimension that complicates

the simple input model: “The second key feature is related to the spatial

dimension in which the military activity takes place: when an organization is

trying to hit a target, all the preliminary steps and the final violent action increase

in complexity with the distance of the target (Polo: 1995, p. 89).” This (relatively

rare) acknowledgement of the spatial dimension of conflict is important and, as

will be noted below, particularly appropriate to the kind of conflict African armies

are now fighting.

The problem with the relatively simple view of military prowess being

related to inputs is that it does not fit the observable facts. In countries where all

institutions are weak, armies cannot simply be assumed to just fight for leaders.

Indeed, national armies are often the most significant threats that leaders face.

6

More generally, whether soldiers will risk their lives for a leader who may or may

not have paid them in the previous few months is a complicated question. Also,

classic questions of morale and esprit de corps—the obsession of many leaders

have even highly institutionalized armies—are always difficult to judge. Thus,

Thom has written, “Whether an army will fight rather than flee when faced with

organized armed opposition is both the simplest yet most difficult judgment to

make. In some instances the answer will differ depending on the nature of the

type of threat and the foe (Thom: 1988, p. 12).” Indeed, warfare in Africa is often

not so much a question of who dominates the battlefield but who runs away first.

For instance, during Laurent Kabila’s march on Kinshasa in 1997, the favored

rebel strategy was not to fight the national army but to devise methods that would

allow the government’s forces to implement their preferred strategy of continua l

retreat. This usually meant that the rebels would not completely encircle the

government forces but would usually leave a corridor for the troops to exit,

however ungracefully. Thus, as the rebels approached a town, the Zairian forces

would “panic, loot, then flee toward the next town where the process would be

repeated.” This pattern characterized most of the “fighting” during the war

although there were pockets of Zairian soldiers who held out and fought. (Thom:

1999, p. 117).

The Evolution of African Militaries

To gain a more nunaced understanding of why African militaries actually

fight, their evolution must be understood. The history of most African militaries

begins at independence in the 1960’s. While the Europeans mobilized large

7

numbers of Africans to fight in each of the two world wars, these forces were

quickly drawn down once those conflicts ended. Most armies during the terminal

colonial period were more recognizable as mechanized police forces than as

combat units. In the British-ruled colonies, even during the terminal colonial

period, when security forces might have been expected to have grown due to the

nationalist upheavals, the number of men in arms was seldom over one per

thousand civilians compared to one in a hundred for Britain and other developed

countries. For both France and Britain, Gutteridge concludes that the low

number of men in arms suggests a, “willingness to operate locally on a very

narrow margin of safety in relation to any security crisis. . . . The European

powers in Africa raised local colonial military forces to fit their world -wide

strategic needs; and their criteria, therefore, rested inevitably on imperial rather

than on local policy. . .(Gutteridge: 1970, p. 316)”

African armies expanded rapidly after independence and took roughly

fifteen years to reach maturity. In 1963, at the dawn of independence, the

average African army had 0.73 soldiers for every thousand people. By 1979, that

figure had more than quadrupled to 3.10 soldiers per thousand citizens. As

economic crisis began to grip the continent, the size of African armies then

began to decrease so that by the mid-1990s, there were only two soldiers per

thousand citizens across the continent (calculated from Morrison, Mitchell and

Naber: 1989, pp. 167-8; U.S. Arms Control and Disarmament Agency: 1985, p.

47 and ACDA: 1995., p. 54). African armies are, by comparative standards,

small. In 1997, African countries (including the relatively well funded countries of

8

North Africa), had on average, only sixty-nine percent as many soldiers, per

thousand citizens, as the average developing country (2.2 versus 3.2) (U.S.

ACDA: 1999, table 1).

Inevitably, the picture of African militaries in general decline becomes

Table 1: Soldiers per thousand citizens, 1987 to 1997 Country Percent Change

1987 to 1997 Country Percent Change

1987 to 1997

Mozambique -82.6087 Sao Tome and Principe -21.5909Ethiopia -74.2857 Swaziland -17.9487Guinea -64.4444 Lesotho -16.6667Nigeria -58.8235 Eritrea -13.9535Equatorial Guinea -50.8475 Chad -12.5Mauritania -50.5495 Malawi -11.1111Ghana -50 Namibia -10.7143Guinea-Bissau -50 Mauritius -10Congo, Rep. -44.9275 Angola -4.25532Senegal -44.4444 Gabon -2.32558Madagascar -40 Kenya 0South Africa -37.931 Mali 0Cameroon -35.7143 Zambia 0Gambia, The -33.3333 Togo 4.166667Sierra Leone -33.3333 Cote d'Ivoire 25Tanzania -29.4118 Sudan 33.33333Niger -28.5714 Uganda 43.75Burkina Faso -27.2727 Botswana 52.94118Congo, Dem. Rep. -26.6667 Benin 55.55556Central African Republic -26.3158 Burundi 225Zimbabwe -24.4898 Rwanda 550Calculated from U.S. Arms Control and Disarmament Agency: 1999. more complicated when the focus is placed on individual countries. Table one

arrays African countries according to the change in the size of their armies over

the last decade and does demonstrate a decline in the armed forces of most

countries. Indeed, most forces did experience a reduction not only in manpower

but also in underlying capabilities. As William C. Thom stated:

Most African state armies are in decline, beset by a combination of

shrinking budgets, international pressures to downsize and demobilize,

9

and the lack of the freely accessible military assistance that characterized

the cold war period. With few exceptions, heavy weapons lie dormant,

equipment is in disrepair, and training is almost nonexistent. . . . the

principal forces of order are in disorder in many countries at a time when

the legitimacy of central governments (and indeed sometimes the state) is

in doubt (Thom: 1995, p. 3).

For instance, a parliamentary report on Zimbabwe’s army—long thought to be

one of the most competent militaries on the continent—found that the force had

only five percent of its vehicles in working order, monthly pilot training had been

abandoned, and seventy percent of the troops in one brigade had been off duty

for a year or more, on forced leaves to save money (Harare Zimbabwe Standard:

1998).

Increasingly, weak or failing states find it difficult to even pay key elements

of their security apparatus, causing a gradual decay in public order. Low or

negative per capita economic growth in many African countries suggests that this

sort of gradual dissolution will become even more common in the future. At the

same time, a few African countries have developed unprecedented abilities to

project force in other countries even while they are unable to control their own

territory. The ability of the Angolan military to move forces in division size

echelons to attack the regime of Patrick Lissouba in Congo-Brazzaville in 1997

was perhaps the best indication of the new capabilities of a small number of

African militaries to operate away from their own borders. Yet, at the same time,

the Angolan military has been unable to defeat the UNITA insurgency in the

10

southern part of its own country. The ability of Uganda, Rwanda, Zimbabwe and

Namibia all to operate forward positions in Democratic Republic of Congo also

suggests that at least a handful of countries can operate in foreign theaters

although the internal defenses of these countries remains problematic. Indeed,

Uganda has become a significant military force in the Democratic Republic of

Congo even while it is unable to defeat insurgents coming into its own territory

from Sudan. As will be explained below, there are good reasons why countries

may decide to gain greater competencies in foreign adventurism than in internal

defense.

Threat Analysis

To better understand how African armies actually fight, and when they will

fight, it is important to understand the threat they face and their ability to respond.

Of course, the actual analysis of the relationship between the deployment of

soldiers, the use of coercion, and victory is, not surprisingly, a very difficult set o f

issues to understand. As Brockett notes, “The consequences of government

repression for mass protest and rebellion have been the subject of much

scholarly attention. Theories have been advanced for linear relationships, but in

both negative and positive directions. Curvilinear relationships have also been

proposed, again with the curves running in both directions. Each of these four

models have found some empirical support—but also contradiction—from a

variety of cross-national aggregate data studies (Brockett: 1995, p. 118).” In

order to get out of this analytic box, it is particularly important to describe the

11

particular circumstances of combat in Africa and the possible strategies of rebels

and of national armies.

The context of most African countries about to experience conflict is that

intelligence of what is happening on the ground is weak. In Africa, even

countries with relatively well-resourced security and intelligence agencies often

do not understand how they have been penetrated by insurgents until after the

fighting has begun. For instance, the presence of cells of the terrorist Osama bin

Laden in Kenya, revealed after the US Embassy in Nairobi was blown-up in

1998, appears to have come as a surprise to the authorities. Similarly, South

African authorities have been generally unsuccessful in understanding the

degree to which foreign Muslim powers have developed ties with PAGAD, a local

vigilante/criminal gang. Countries with fewer intelligence resources often have

little idea that conflict is about to begin within their own territory.

Initial phases of insurgency

While every insurgency, like every unhappy family, is different, some

generalizations about the threat facing national armies across Africa can be

made. Insurgencies generally start small. The eleven men who started the

fighting in Eritrea (Pateman: 1998, p. 117), the famous twenty-seven fighters

who began the National Resistance Movement’s campaign in Uganda

(Museveni: 1986, p. 7), the approximately one hundred soldiers of the National

Patriotic Front that crossed into Liberia with Charles Taylor (Ellis: 1999, p. 110),

the thirty-five trained soldiers who started the RUF in Sierra Leone (Abdullah and

Muana: 1998, p. 177), the two hundred and fifty that started FRELIMO

12

(FRELIMO: 1982, p. 147) are representative of how vulnerable and small rebel

movements are at the beginning. Of course, many rebel movements are

defeated early or simply collapse from their internal divisions, never to be heard

from again. As a result, rebel leaders are, literally from the start, acutely

conscious of the coercive power of the state and opportunities to seek refuge

from the state.

It is at this early stage, of course, that an insurgency is most likely to be

defeated. Brockett concludes from Central America: “during ‘normal conditions’

that is, prior to the onset of a protest cycle, escalating repression will deter

popular mobilization against the regime. In contrast, in the ascendant phase of

the protest cycle, the same repression is likely to provoke increased mass

oppositional activities (1995, p. 134).” It is not surprising that when the cost of

repression fall heavily on members of a small rebel movement that they are

individually deterred (if not killed outright) but that governments find it impossible

to keep the cost of repression for each individual from falling as insurgent group

size begins to grow rapidly.

For African militaries, even rebellions in the relatively vulnerable early

stages are difficult to defeat. Given how weak intelligence systems are and how

small the insurgencies are at the beginning, it is necessary to mobilize a

significant amount of firepower in a short period of time to fight rebels when they

are most vulnerable. Indeed, militaries with intelligence services far more

capable than African states routinely mistake insurgents for criminals in the early

days of a rebellion, thereby delaying the implementation of a counter-insurgency

13

campaign (Rich and Stubbs: 1997, p. 7). African militaries are far less capable

than the national forces in Central America and they often operate over much

larger territories. The very smallness of insurgencies also makes them difficult to

detect and defeat and African countries are unlikely to have the necessary

intelligence assets and co-ordinated police/military capabilities to undertake

surgical strikes.

As a result, African militaries rely on relatively blunt strikes in the hope of

defeating even small groups. Also, given that the basic local political structures

of most African counties are so underdeveloped, they usually cannot marry force

projection with a nuanced political strategy: they have no carrots, only sticks.

Indeed, the state finds any kind of local vision that might emerge from even a

proto-insurgency movement to be extremely threatening (Watts: 1997, p. 36).

Thus the following description by Watts of Nigeria’s reaction to the Prophet

Maitatsine and his followers in Kano in 1980 is typical of what African states

would at least like to do to small groups of insurgents,

Federal Nigerian armed forces began a massive military assault on the

sleepy residential and commercial quarter of ‘Yan Awaki in the walled city

of Kano, in Northern Nigeria. Under ferocious aerial and ground

bombardment, somewhere between five and ten thousand people were

slaughtered, and another fifteen thousand injured. The body of the self-

proclaimed leader, Maitatsine . . . who was killed in the conflagration, was

exhumed by the military authorities from a shallow grave outside the city

walls where he had been laid to rest by his followers, and placed on

14

display at the local police. . . Those federal authorities who engaged in

the bloodletting and who self-consciously sought to quite literally

obliterate a leader and a community which saw itself as incontrovertibly

Muslim in character and constitution, were acting on the instructions of a

civilian administration presided by an aristocratic northern Muslim

(President Shagari) and a government that was widely held to be

dominated by northern Muslim elites (the so-called “Kaduna Mafia”)

(Watts: 1997, pp. 36-7).

In fact, very few African armies would have been able to duplicate the firepower

utilized by the Nigerians against this minor religious group, although few would

hesitate if they could. More often, they attempt to use overwhelming force but

fail, often with disastrous consequences. For instance, in reaction to the

incursion of Charles Taylor’s one hundred guerrillas of the National Patriotic

Front of Liberia (NPFL) in Nimba County in 1989,

Government forces in the area had responded to the first NPFL attack by

detaining suspects, which in practice meant singling out young Gio men

for arrest or murder while committing other brutalities in a region which the

Krahn—dominated milita ry regarded as enemy territory. When the

government imposed a curfew throughout Nimba Country, thousands of

local people fled into Guinea and Côte d’Ivoire, leaving Armed Forces of

Liberia (AFL) troops to loot the empty towns they found. AFL counter-

measures were ineffective since the army was receiving no support from

any part of the population in Nimba County other than its Mandingo

15

residents, many of them traders and shopkeepers. Substantial numbers

of government soldiers were deserting. . . (Ellis: 1999, pp. 77-8).

Mature insurgencies

If the small rebel group survives this initial onslaught, the government

must fight a more classic type of counter-insurgency war. As hard as it is for

African militaries to fight small gangs of insurgents, more widespread warfare

involving larger echelons of troops is even more difficult. As Thom once again

notes,

Counterinsurgency is a most difficult job for any military to perform, and it

requires African armies to perform in areas where they are the weakest;

i.e., aggressive, multiple field operations with little short term payoff that

must be sustained over long periods to achieve the ultimate goal. This

requires good logistic support, mobility, maintenance, training in small unit

tactics, good leadership, especially at the junior officer and NCO levels,

and an ability to integrate military intelligence with field operations (Thom:

1988, p. 7).

In addition, counter-insurgency places significant demands on a government to

co-ordinate the (very difficult) military tasks with a comprehensive political

strategy. The United States’ own doctrinal statement on low-intensity conflict

suggests just how difficult developing a comprehensive political/military counter-

insurgency campaign can be:

(Internal defense and development) focuses on building viable political,

economic, military, and social institutions that respond to the needs of

16

society. Its fundamental goal is to prevent insurgency by forestalling and

defeating the threat insurgent organizations impose and by working to

correct conditions that prompt violence. . . The government often must

overcome the inertia and incompetence of its own political system before it

can cope with the insurgency against this system. This may involve the

adoption of reforms during times of crisis, when pressures limit flexibility

and make implementation difficult (Departments of the Army and Air

Force: 1990, pp. 9 -10).

Rebels generally have more choices as an insurgency proceeds. Indeed,

“It is relatively easy in military terms to operate successfully as a guerrilla force

against a black African state: strategic soft targets such as rail lines, bridges, oil

pipelines, power stations and transmission lines are difficult to protect, yet their

destruction has great economic and psychological impact (Thom: 1988, p. 8).”

This mode of attack allows guerillas to strategically retreat from the armed forces

of the state rather than have to fight, making counter-insurgency all the more

difficult. As Yoweri Museveni noted during his struggle for power in Uganda,

“Loss of territory is, at this stage, of no consequence. In our case, the more

important considerations are the preservation and expansion of our forces by

avoiding unnecessary casualties, and destroying the enemy’s means of making

war. . . (Museveni: 1986, p. 12).” As the conflict cycle proceeds, the tasks of

rebels become relatively easier while the government finds it increasingly difficult

to fight.

17

The utility of African states’ limited military assets therefore begins to

decline as

Figure 1 about here

the insurgency continues. Indeed, the armies in Africa that have waged even

temporarily successful military campaigns have had to make extraordinary efforts

that go well beyond the capabilities of most national armies. For instance, the

Portuguese response to guerrilla mines in Angola and Mozambique was a

massive road tarring exercise that exceeded even what the British attempted in

Malaya or the US in Vietnam. More generally, the entire Portuguese Army and

Air Force was reconfigured to fight guerrillas (Beckett: 1985, pp. 145-6). Even

then, while the Portuguese army performed relatively well in Mozambique and

Angola for several years, especially considering the logistical task of operating so

far from Lisbon, eventually the colonizer lost the political struggle and left Africa

in defeat.

After the initial flurry of combat that reflects the government’s initial

attempt to obliterate the rebels and the insurgents’ desire to inflict a

psychological wound on the army, rebels often establish a base in the

countryside to organize a mobile force and recruit new soldiers. This was the

pattern of insurgents in Rwanda, Uganda and the Popular Movement for

Salvation in Chad. After logistical lines are consolidated, a push for the capital is

made. Some governments are so weak that they cannot defeat even this

lightening strike and the capital is overrun. Small countries are particularly

18

vulnerable to this type of two-phase campaign (see generally, Young: 1996, p.

185).

Of course, many insurgents are not able to adopt a quick capital-first

strategy and get mired in the countryside. This may occur because of the

government is able to provide point defense of the major city or because the

territory is so big that the rebel’s base is far enough from the capital that their

limited logistical ability cannot triumph over unyielding geography (e.g., Angola,

Sudan). At this point, it is possible that a stalemate begins that can continue for

many years. On the hand, the rebels cannot defeat the government because

defeat is defined by the international community as militarily occupying the

capital (See, generally, Herbst: 2000b). On the other hand, the military

perquisites for fighting a protracted counter-insurgency campaign are simply

beyond the capabilities of most African militaries.

As a result, African states based in capitals and guerrillas anchored in

relatively large countrysides often reach a kind of equilibrium. The government

knows that it is too costly to attack the guerillas in order to defeat them and

therefore can define a territory that it can defend. It will be politically viable as

long as it controls the capital and some other towns. For instance, the

Rhodesian notion of “vital area ground--” territory whose capture or control would

result in or significantly contribute to national defeat--was developed in the

terminal stages of that national liberation struggle in what is now Zimbabwe.

Specification of what territory Salisbury had to control allowed a significant

portion of the countryside to be ceded to guerrillas. In fact, as figure two (from

19

the Rhodesian central military authorities) demonstrates, a country’s leadership

in the midst of a brutal war can tolerate

Figure 2 about here

significant losses of territory without conceding defeat. However, the Rhodesian

strategy, while preventing outright military defeat, did not allow for victory

because small groups of guerrillas could easily operate in territory that was not

considered vital and thereby gain popular allegiance (Cilliers: 1985, p. 249).

Similarly, the French, as figure three suggests, tolerated a significant erosion of

their territory in Algeria before finally conceding defeat. Many other

Figure 3 about here

African governments have at various times lost significant control over their

hinterlands but remained in power because they had physical control of the state.

Thus, President Mobutu of Zaire was really little more than the mayor of

Kinshasa in his declining years while the governments of Sudan, Angola, and

Mozambique, among others, have lost control over significant parts of their

territory over many years although were not militarily defeated.

The model presented here directly contradicts significant assumptions of

some of the political economy models of conflict. For instance, in Hirshleifer’s

model that military forces are used when the decisiveness of conflict is

“sufficiently great ((1991, p. 197).” The problem with that model is that it does

not recognize the life cycle of insurgency and counter-insurgency. Military force

may only be sufficiently great to actually defeat insurgencies in Africa at the very

beginning of the conflict but that will be precisely the time that many armies will

20

be unable to deploy. Hirshleifer assumes that military assets are always

available and can be deployed at any time. Such an assumption is unrealistic for

conflicts in Africa. More generally, the “blackbox” model of relating military

prowess to inputs simply fails to take account the growing disutility of military

forces as insurgencies progress.

The Ability to Mobilize

The ability of an African army to defeat rebels therefore greatly depends

on its ability to mobilize forces relatively quickly in the face of a growing enemy

threat. If it is able to mobilize and deploy the agents of violence, there is some

chance that it can crush an insurgency while the rebels are still particularly

vulnerable. However, as time goes on, it becomes increasingly difficult to fight

an insurgency and states therefore become vulnerable to outright defeat

(especially if they have a relatively small territorial base and face relatively adept

rebel fighters) or a prolonged war which neither side can win but which will

almost surely devastate the country. Indeed, a particularly important reason for

militaries to defeat insurgencies early on is so that leaders do not simply become

accustomed to the idea of becoming “stationary bandits” over smaller and

smaller pieces of their own territory and not even be interested in gaining control

over their own territory. The international community will tolerate this type of

shirking of responsibility but it is usually disastrous for the country.

Table 2 presents some instances of unsuccessful mobilization in the face

of an enemy while table 3 presents instances where African countries were able

to mobilize in the face of a growing enemy threat. I have used soldiers per

21

thousand citizens as my major indicator of mobilization because budgetary data

on African militaries are completely unreliable and it is at least possible to

present a reasonable count of the number of soldiers in an army. As much as

possible, I have tried to have

Table 2: Unsuccessful Mobilizations During Conflict

Country Mobilization during crisis (soldiers/thousand

citizens) Congo-Brazzaville 1993 1994 1995 1996 1997 4.2 4.1 4 3.9 3.8 Guinea 1987 1989 1991 1993 1995 1997 4.5 2.7 2.4 2.2 1.7 1.6Liberia 1986 1987 1988 1989 1990 1991 2.6 2.5 2.8 2.7 3.5 2.9Sierra Leone 1987 1989 1991 1993 1995 1997 1.5 1 1.1 1.2 1.1 1Somalia 1982 1984 1986 1988 1990 9.2 6.9 7.4 6.75.6 Sudan 1980 1981 1982 1983 1984 1985 3.4 4.5 4.3 4 2.9 2.8Uganda 1982 1983 1984 1985 1986 1987 0.7 0.9 1 1 1 0.9Zaire 1973 1975 1977 1979 1981 1983 2.8 2.2 2 0.8 1.5 1.3Zaire 1993 1994 1995 1996 1997 1 .9 .9 1.1

Source: Calculated from U.S. Arms Control and Disarmament Agency, various years.

22

Table 3: Successful Mobilizations During Times of Conflict

Country Mobilization during crisis Angola 1979 1981 1983 1985 1987 1989 7.2 7.6 7.4 8.7 9.4 12.9Ethiopia 1974 1976 1978 1980 1982 1984 1.3 1.8 6.1 6.5 6 4.5Mozambique 1979 1981 1983 1985 1987 1989 2.5 2.4 2.4 2.8 4.6 4.6Rwanda 1990 1991 1992 1993 1994 1995 0.8 4.1 4 3.9 6 5.5Somalia 1973 1975 1977 1979 1981 1983 8.3 9.4 16.1 15.4 9.2 7.7

Source: Calculated from U.S. Arms Control and Disarmament Agency, various years.

column five be the year in which the enemy threat became most obvious,

although some of the conflicts do not lend themselves to such neat chronologies.

For instance, 1990 is the year that Taylor invaded Liberia, 1986 is the year

Museveni toppled the government in Kampala, 1997 is the year Angola invaded

Congo-Brazzaville while 1994 is the year that the RFP invaded Rwanda. In other

conflicts, it is more difficult to cite a precise time as the conflicts emerged slowly

and then only gradually built to a crescendo. For instance, the threat to both

Sierra Leone and Guinea grew gradually during the 1990’s as instability from

Liberia spread westward. FRELIMO’s war with RENAMO in Mozambique and

UNITA’s battle with the MPLA government in Angola also gradually became

worse throughout the 1980’s. Thus, while the ability (failure) to mobilize is not as

clear in countries where the war came gradually, it is still possible to judge if they

were able to increase at least the size of their military in the face of a growing

enemy threat. It is notable that of the cases of failed mobilization, most of the wars had a

distinctly civil nature although, as in almost any civil war, there was a significant

international element. Thus, while Congo-Brazzaville was invaded by Angola,

Luanda was working in support of former President Denis Sassou-Nguesso who

23

had worked for years to destabilize his successor. Similarly, the wars in Liberia

and Sierra Leone are domestic conflicts, albeit fed by outsiders. Somalia in the

1990’s, Sudan and Uganda are civil wars classically understood. The failure to

mobilize is quite remarkable. For instance, Siad Barre’s army in Somalia actually

shrunk significantly as his country slid into civil war while the authorities in

Kampala were unable to increase their army at all in the endgame before

Museveni took over. The invasions of Zaire during the Shaba crises of 1977 and

1978 and by Ugandan, Rwandan and Angolan forces in support of Laurent

Kabila in 1996 and 1997 are the clearest examples of a foreign invasion which

was not met by significant mobilization.

The cases of successful mobilization are more clearly linked to external

invasion although some of the cases cannot be parsed cleanly. Somalia

mobilized to take what it viewed as the Somali portion of Ethiopia in 1977-78

while Ethiopia mobilized slightly later to defend itself. Similarly, the Hutu

authorities in Rwanda did mobilize in anticipation of the invasion by the RPF in

1994. Angola and Mozambique are more clearly civil wars but both Maputo and

Luanda had enemies (RENAMO and UNITA) that were directly tied to South

Africa. They could thus legitimately claim during at least pleas for mobilization

that they were under threat from foreign powers.

Why would governments have an easier time mobilizing in the face of

foreign threats than conflicts that are more internal in nature? Why do some

armies seem to operate well across borders while unable to defeat rebels in their

own country? At least part of the answer is that purely domestic conflicts within

24

African countries are a reflection of basic failures in institutions and policies.

Indeed, many failing African states cannot mobilize in the face of insurgency

because they do not have the resources or the capabilities to maintain even the

agents of violence in their country, despite the fact that the monopoly on

legitimate violence is the defining characteristic of the state. However, when a

state is attacked by another state, it may be because of reasons within the

invader and not because the target is necessarily dysfunctional. Thus, Somalia

attacked Ethiopia because it wanted to be reunited with the Somalis in the

neighboring countries not because of some fundamental institutional breakdown

in Ethiopia. As a result, Ethiopia was able to mobilize in the face of Somalia’s

attack and was able to defeat the enemy. Similarly, Ethiopia attacked Eritrea in

1998 not because Eritrea was failing but because of a complex set of commercial

and political conflicts between the two countries. Eritrea was therefore able to

mobilize against the Ethiopia attack although it did suffer a significant defeat.

Of course, it may also be easier to mobilize the population in the face of

an external threat. Indeed, the presence of a palpable external threat may be the

strongest way to generate a common association between the state and the

population. External threats have such a powerful effect on nationalism because

people realize in a profound manner that they are under threat because of who

they are as a nation; they are forced to recognize that it is only as a nation that

they can successfully defeat the threat. Anthony Giddens recounts the effects of

World War I: "The War canalized the development of states' sovereignty, tying

this to citizenship and to nationalism in such a profound way that any other

25

scenario [of how the international system would be ordered] subsequently came

to appear as little more than idle fantasy (Giddens: 1987, p. 235)." Michael

Howard notes the visceral impact of wars on the development of nationalism

throughout Europe:

Self-identification as a Nation implies almost by definition alienation from

other communities, and the most memorable incidents in the

group-memory consisted in conflict with and triumph over other

communities. France was Marengo, Austerlitz and Jena: military triumph

set the seal on the new-found national consciousness. Britain was

Trafalgar—but it had been a nation for four hundred years, since those

earlier battles Crecy and Agincourt. Russia was the triumph of 1812.

Germany was Gravelotte and Sedan (Howard: 1978, p. 9. Emphasis in

the original).

Inevitably, it is much more difficult to mobilize a population during a civil war

given that the conflict itself reflects fundamental dissatisfactions with the existing

political order.

Finally, African countries may be able to mobilize to attack other countries,

even while the security of their existing homeland is not guaranteed, simply

because the territory in other countries may be more valuable than their own

homelands. As noted above, large countries especially, can afford to have

insurgents control significant amounts of territory as long as they did not threaten

the capital. At the same time, territory in adjoining countries, or the policies of

those neighbors, may be more important to control or influence. For instance,

26

the government in Luanda probably felt it more important to intervene to

guarantee the course of events in Brazzaville—itself relatively close to Luanda—

even while the rebellion in the far south of Angola continued. Similarly, Uganda

seems to have thought it more important to try to change security arrangements

in eastern Democratic Republic of the Congo than in its own north, probably

because events in eastern Congo posed a more significant threat to Kampala.

There have also been repeated reports—almost all unconfirmed—that Uganda

and other countries operating in DRC have been able to recoup some of their

expenses through theft of natural resources. While such cost recovery is much

harder than most people believe—running a mine is no guarantee of profit in the

Congo, as many other failed businesses have discovered—it may very well be

the case that controlling some territory of adjoining countries is potentially more

lucrative than defense of the homeland.

Again, models which try to associate the input of resources with military

prowess will not necessary predict correctly how African countries will react.

Mobilization in the face of an enemy is difficult and the ability to convince people

that they should fight and die for their country may depend on the enemy. At the

same time, leaders may make very different decisions on what they are going to

fight for given calculations about the relative value of particular pieces of territory

within or without their homeland. Two different types of threat—from internal civil

war and external enemies—confront African leaders and they may have very

different options according to what kind of war they are involved in.

27

Conclusion

This analysis of African militaries suggests that wars in Africa are far

harder to end than models which relate inputs to military prowess suggest.

Perhaps the only time when African armies—given their precarious state and

their difficulties in mobilizing to face domestic conflict—may have a good chance

of defeating an insurgency is at the start of hostilities. Of course, that is precisely

when African militaries are least prepared to fight. It is also important for

insurgencies to be defeated early because leaders may simply fight it to their

advantage to withdraw from territory and rule over less rather than fight hard as

the insurgency continues.

From a policy perspective, these findings may suggest that the optimal

level of investment in the military and security forces is higher than usually

believed. If it is essentially too late to invest in the military once fighting has

begun in earnest—at least in domestic conflicts—African countries may have to

have invest in more competent standing forces which could be used at the start

of an insurgency. However, such an investment in preparedness would also

mean that funds would often be used to ready an army that, as it turns out, would

not have an enemy to fight. The opportunity cost of such an investment would

therefore be quite high.

28

Figure 1: Effectiveness in Combat High Effectiveness in combat Low Time

29

Figure 2: Zimbabwe: Vital Area Ground

Source: Cilliers: 1985, p. 251

30

References

Abdullah, Ibrahim and Patrick Muana (1998), “The Revolutionary Front of Sierra Leone,” in Christopher Clapham (ed.) African Guerillas. London: James Currey. Beckett, Ian F.W. (1985), “The Portuguese Army: The Campaign in Mozambique,” in Ian F.W. Beckett and John Pimlott (eds.), Armed Forces and Modern Counter-Insurgency. NY: St. Martin’s Press. Brockett, Charles D. (1995), “A Protest-Cycle Resolution of the Repression/Popular-Protest Paradox,” in Mark Traugott (ed.), Repertoires and Cycles of Collective Action. Durham: Duke University Press. Charap, Joshua and Christina Harm (1999), “Institutionalized Corruption and the Kleptocratic State,” International Monetary Fund Working Paper WP/99/91. Cilliers, J.K. (1985), Counter Insurgency in Rhodesia. London: Croom Helm. Clapham, Christopher (1998), African Guerillas. Oxford: James Currey. Collier, Paul and Anke Hoeffler (1998), “On Economic Causes of Civil War,” Oxford Economic Papers 50. Departments of the Army and Air Force (1990), FM100-20 and AFP 3-20: Military Operations in Low Intensity Conflicts (Washington, DC: Department of Defense). Ellis, Stephen (1999), The Mask of Anarchy: The Destruction of Liberia and the Religious Dimension of an African Civil War. NY: New York University Press. FRELIMO (1982), “On the Necessity of Prolonged War,” in Aquino de Bragança and Immanuel Wallerstein (eds.), The Strategy of National Liberation. London: Zed Press. Giddens, Anthony (1987), A Contemporary Critique of Historical Materialism, vol. 2 . Berkeley: University of California Press, 1987. Grossman, Herschel I. (1995), “Insurrections,” in Keith Hartley and Todd Sandler (eds.), Handbook of Defense Economics. Amsterdam: Elsevier. Gutteridge, William F. (1970), “Military and Police Forces in Colonial Africa,” in L.H. Gann and Peter Duignan (eds.), Colonialism in Africa, 1870-1960, vol. 2. Cambridge: Cambridge University Press.

31

Harare Zimbabwe Standard (1998), “Zimbabwe: Report cites UK Paper on “Ominous” State of Defence Force” May 11th edition reproduced in Foreign Broadcast Information Service, 12 May 1998).

Heggoy, Alf Andrew (1972), Insurgency and Counterinsurgency in Algeria. Bloomington: Indiana University Press. Herbst, Jeffrey (2000a), “The Organization of Rebellion in Africa,” presented at the conference on “The Economics of Political Violence,” sponsored by the Center of International Studies and the World Bank Research Group, Princeton University,18-19 March 2000. Herbst, Jeffrey (2000b), States and Power in Africa: Comparative Lessons in Authority and Control. Princeton: Princeton University Press. Hirshleifer, Jack (1988), “The Analytics of Continuing Conflict,” Synthese 76 (August). Hirshleifer, Jack (1991), “The Paradox of Power,” Economics and Politics 3 (November). Howard, Michael (1978), War and the Nation State. Oxford: Clarendon Press, 1978. Morrison, Donald George, Robert Cameron Mitchell, and John Naber (1989), Black Africa: A Comparative Handbook, second edition. NY: Paragon House. Museveni, Yoweri (1986), Selected Articles on the Uganda Resistance War. Kampala: NRM Publication. Olson, Mancur (2000), Power and Prosperity: Outgrowing Communist and Capitalist Dictatorships. NY: Basic Books. Pateman, Roy (1998), Eritrea: Even the Stones are Burning . Revised edition. Lawrenceville [NJ]: Red Sea Press. Polo, Michele (1995), “Internal Cohesion and Competition among Criminal Organizations,” in Gianluca Fiorentini and Sam Peltzman (eds.), The Economics of Organized Crime. Cambridge [UK]: Cambridge University Press. Reno, William (2000), “Shadow States and the Political Economy of Civil Wars,” in Mats Berdal and David M. Malone (eds.), Greed and Grievance: Economic Agendas in Civil Wars. Boulder: Lynne Rienner. Rich, Paul B. and Richard Stubbs, “Introduction: The Counter-Insurgent State,” in Paul B. Rich and Richard Stubs (eds.), The Counter-Insurgent State: Guerrilla Warfare and State-Building in the Twentieth Century. London: Great Britain.

32

Russell, D.E.H., (1974), Rebellion, Revolution, and Armed Force: A Comparative Study of Fifteen Countries with Special Emphasis on Cuba and South Africa. NY: Academic Press. Snyder, Charles P. (1986), “African Ground Forces,” in Bruce E. Arlinghaus and Pauline H. Baker (eds.), African Armies: Evolution and Capabilities. Boulder: Westview. Thom, William G. (1988), “Judging Africa’s Military Capabilities,” mimeo. Thom, William G. (1995), “An Assessment of Prospects for Ending Domestic Military Conflict in Sub-Saharan Africa,” CSIS Africa Notes, 177 (1995).

Thom, William G. (1999), “Congo-Zaire’s 1996-97 Civil War in the Context of Evolving Patters of Military Conflict in Africa in the Era of Independence,” The Journal of Conflict Studies 19 (Fall).

U.S. Arms Control and Disarmament Agency (1985), World Military Expenditures and Arms Transfers 1985. Washington, DC: ACDA.

U.S. Arms Control and Disarmament Agency (1995), World Military Expenditures and Arms Transfers 1995. Washington, DC: ACDA.

U.S. Arms Control and Disarmament Agency (1999), World Military Expenditures and Arms Transfers 1999. Washington, DC: ACDA.

Watts, Michael (1997), “Black Gold, White Heat: State Violence, Local Resistance and the National Question in Nigeria,” in Steve Pile and Michael Keith (eds.), Geographies of Resistance. London: Routledge. Young, Eric T. (1996), “The Victors and the Vanquished: The Role of Military Factors in the Outcome of Modern African Insurgencies,” Small Wars and Insurgencies 7 (Autumn). Young, Tom (1997), “A Victim of Modernity? Explaining the War in Mozambique,” in Paul B. Rich and Richard Stubbs (eds.), The Counter-Insurgent State: Guerrilla Warfare and State Building in the Twentieth Century. London: Macmillan.