INTRODUCTION

In the 1970s and 1980s, Africa was the most militarised and conflict-torn continent in the world. Civil wars, military coups, armed foreign interventions, ethnic conflicts, and other forms of civil and political strife were commonplace. At the height of these conflicts one-quarter to one-third of Africa's population were displaced and became refugees. In the 1970s, military budgets and arms procurements in Africa grew at a faster rate than in any other world region, before receding in the 1980s under the impact of deepening economic recession. While these trends have not disappeared altogether, Africa has witnessed a no less dramatic demilitarisation and democratisation of its politics since the late eighties and early nineties. Peace accords and elections have led to the end of civil wars in Namibia, South Africa, Mozambique, Ethiopia and Eritrea. Military governments have been removed in Mali, the Central African Republic, Benin, the Congo, Ghana, Mauritania, Burkina Faso and, until recently, Niger and Burundi. Authoritarian and one-party governments in the Ivory Coast, Kenya, Gabon, Cameroon, Cape Verde, Tunisia, Zambia, Malawi, Tanzania and elsewhere have submitted themselves to elections and otherwise liberalised their regimes.

THE PATTERNS OF MILITARISATION AND DEMILITARISATION

However, the patterns of militarisation and demilitarisation on the continent are both highly nuanced (in analytical terms) and require some definition and comment before discussing their implications. While the militarisation of the African continent has been both extensive and endemic, its nature, depth and relative distribution have been quite uneven and defy easy generalisation.

Civil-military relations have followed distinct regional and national trajectories. Coups have been less prevalent in Southern Africa, but common in West, Central and Eastern Africa. While some African countries have been subject to persistent military coups, others have escaped the coup contagion altogether. Part of the explanation derives from the pattern of decolonisation. On the whole, regimes that came to power through armed struggle were able to create or reorganise their own armed forces, and were thus more successful in imposing control than those that did not. Secondly, among this second set of countries, some, such as Tanzania, Kenya and Senegal, following initial challenge from their armed forces, were able to reorganise and institutionalise their civil-military relations and to sustain civil rule, while others, such as Ghana and Uganda, were unable to do so. Similarly, certain military regimes were able to terminate, or successfully avoid military interventionism and create durable political dispensations, while others succumbed in turn to a further cycle of coups.

The size of military and defence spending varies quite substantially between individual African countries. The data suggests that a small number of countries control a disproportionate amount of the accumulated and current military resources on the continent. Further, the acquisition of military resources bore little consistent relation to economic means in the past. One of the peculiarities in the African case is the fact that with few exceptions, processes of militarisation have actually tended to be exacerbated among the poorest African societies (Mozambique, Somalia, Ethiopia, Chad, Uganda and – in terms of military spending – Tanzania), again verifying the observation that war has been the most important impulse underlying militarisation in Africa. Even here there are historical variations: as seen earlier, military spending and arms acquisitions rose sharply overall from the mid-1970s, and contracted as sharply from the mid-1980s.

The militarisation of politics is not limited to situations of overt military intrusion into politics. For instance, while the South African Government remained formally civilian, both the politics and economy of South Africa became heavily militarised, with the military and security and intelligence agencies exercising more institutionalised influence than in most so-called 'military' regimes. As is well known, some of the most durable civilian regimes, such as Kenya or Cameroon, have been (and continue to be) particularly notorious for repressive and illegal use of military and special security units. Civil-military relations under such civilian regimes have tended to be highly personalised and regime-specific (the transition from Ahidjo to Baya in Cameroon, and on a different level Sekou Toure in Guinea, suggests that there is no guarantee that the new regime can simply step into the shoes of its predecessor). This bears out the frequent observation that civilian control is not in itself a necessary guarantee of democracy. The need to demilitarise the politics of such regimes is at least as urgent as that of the military dictatorships.

One should also grasp the historicity of the problem of militarisation: The earlier problematic of the 1960s and '70s – as to whether governments (both civil and military) could maintain control over their military forces – was increasingly displaced by the question of whether the state could maintain its monopoly over the instruments of violence. The process of diffusion of the means of violence away from the state and official armies was the most significant development in the process of militarisation on the African continent in the 1980s and '90s. First, official militaries splintered, challenged from within by factions or from below by their own subaltern ranks; then guerrilla and other informal armed bodies arose to contest the state's monopoly of force. This 'democratisation' of the instruments of violence (which had occurred earlier in the 'liberation territories') undermined both the state and the standing army, intensifying the interlocking crisis of civil and military authority. Indeed in some countries, the state and the military hierarchy have so completely lost control over the legitimate monopoly and central control of coercion, that it is moot whether the main threat of organised violence comes from the state and official military command or from other informal entrepreneurs of violence.

In countries such as Somalia and Liberia, militarisation has a generational dimension with disturbing implications for the future and the reproduction of the culture of violence. In these countries the militarisation of children and the youth is the most eloquent testimony of the collapse of the official monopoly of force and of the extent to which violence has taken over the moral fabric of African society. Children have been doubly victimised by war, both as victims of and dealers in violence. Child soldiers have featured in most civil wars since the Biafran war. Warlords, in particular, have discovered that children make ruthless and efficient killers. Children were also at the receiving end of apartheid and destabilisation in Southern Africa, and now of ethnocide in Rwanda, Burundi and Liberia. With the militarisation of culture and the collapse of the mechanisms of socio-economic reproduction in some of these countries, the gun has become the tool of opportunity for youth.

Militarisation and war also carry serious implications for gender structures and equality. The military is the most gender-imbalanced of all major social institutions, the embodiment par excellence of masculine-patriarchal power. National security doctrines cast women (with children) as helpless dependants constitutionally incapable of the military arts and self-defence, and therefore requiring protection by men. In contradiction, the military ethic has regarded women through the ages as legitimate war booty. Rape and abuse of women has thus risen dramatically with the militarisation of African society. War has disrupted what little progress women have made on the margins of African society and deepened their marginalisation. At another level, military regimes have exacerbated the gender cleavage and the political disempowerment of women; at the same time these regimes have sought to compensate with transparent schemes of patriarchal mobilisation of women. It is clear from this that the question of demilitarisation is of the most dire consequence for women in particular.

Given the proliferation of armed formations in Africa, the meaning of the term 'military' is no longer as apparent as one used to assume. The components of the military system vary from country to country and express particular combinations of armed forces, gendarmerie, intelligence, police, secret police, customs and other security agencies. Official militaries have also tended to be recomposed over time by succeeding civil and military administrations, institutionally as well as politically and ideologically. More complex, in much of Africa traditional concepts of the 'military' are being challenged by the great variety of formal and informal organisational formats (guerrilla armies, warlord formations, armed gangs and death squads, proxy forces of all kinds) under which Africans have taken up arms, frequently successfully, and pursued territorial and non-territorial ambitions. In some countries, 'demilitarisation' involves the demobilisation and/or integration and political emasculation of competing official and non-official armed bodies and gangs (e.g. South Africa). In others demilitarisatioon refers primarily to the national armed and security services.

The movement toward demilitarisation has shown a similar unevenness, both as process and outcomes. In one set of cases (Benin, Mali, the Congo, Central African Republic, initially Burundi and Niger, and – on a different plane – South Africa), the military has acquiesced to complete withdrawal and the emergence of an elected democratic system. Closer examination of these cases reveals further divergences in the actual conduct of the military during these transitions, ranging from the neutrality of the Benin's armed forces, to the active intervention of the Malian armed forces on behalf of pro-democracy forces, to the hostility displayed by armies in Niger, the Congo and Burundi toward democracy movements. In a second set of cases (Ghana, Burkina Faso, Guinea, Mauritania, and lately Uganda), the military collaborated directly or indirectly in regime rearrangement (liberalisation) and the civilianisation of the existing authoritarian leadership in closely controlled transition processes. In yet a third case (Nigeria and Algeria), the military initiated and then aborted transitions at relatively late stages of the process when the 'wrong' political party appeared poised to win the elections, while in the fourth and final set of cases (Togo and Zaire), the military and security agencies provided the backbone for successful regime resistance and deflection of democratic pressures.

Equally divergent have been the actions of national militaries following the initial transition and installation of democratically elected authorities. Political actions by the military following initially successful transitions, have ranged from complete displacement of the new regime (Niger, 1996), to a 'surgical' strike designed to destabilise and incapacitate the regime and halt further democratisation, initially without supplanting the regime altogether (Burundi, October 1993),2 as well as series of 'trade union' actions linked to pay and other disputes (Lesotho in January and April 1994, Guinea in February 1996, Central African Republic in April 1996). In these instances, the military has withdrawn (or regrouped) formally rather than substantively, launching overt or covert actions to pressure the regime publicly or from behind the scenes to limit further democratisation, prevent encroachment on its prerogatives, address grievances, and so on.

This uneven pattern of military responses seems to reflect specific corporate and contextual factors:

the political and ideological calculus of particular military formations;



the outcome of political contests within the military itself and the attitudes of the dominant factions emerging from these struggles;



the decomposition and increasing fracturing of the military and security agencies that constituted the backbone of the authoritarian regime;



the intensity of pressure from foreign patrons, leading to the precipitous collapse or reconfiguring of authoritarian regimes;



the degree of professionalism and autonomy from social forces, which may make it possible or difficult for the military to disentangle its corporate interests from those of the regime and particular ethnic, regional, or social interests; as well as



the nature of the transition itself.

In sum, militaries do not have the same dynamics, orientations, or motivations. This requires that the discussion be grounded in understanding specific forms of militarisation, and the particular problems of transition/ demilitarisation to which they give rise in each national case.

It is also undeniable that the forms of 'democracy' that have emerged on the continent differ greatly in quality and depth. Considerable caution is therefore required in assessing their significance. Firstly, in few of these countries (Benin, Zambia, Malawi, Mali and South Africa among them) elections have actually unseated incumbent governments or shifted power decisively from the pre-existing political class. In a large number of cases elections have merely 'constitutionalised' existing authoritarian regimes, military as well as civil, with former dictators donning a thin mantle of democracy. In formerly single-party states such as Kenya, the Ivory Coast, Gabon, and Cameroon, the official party has re-emerged as the dominant party within the framework of political pluralism; while in former military regimes in Ghana, Burkina Faso, and Mauritania a parti militaire emerged to exploit the advantages of incumbency, winning elections of which the rules were set, by and large, by the governing regime. While several democratic freedoms have been realised (mainly in the area of the press and the independence of the judiciary), these are still fragile and circumscribed, and depend on the continued goodwill of the political leadership. In several of these transitions (Kenya is again a good example), after a brief period of liberalisation, the regime has returned to its old repressive style of rulership. Secondly, important pillars of the authoritarian structures that undergird African states continue to survive irrespective of the nature of the political leadership, particularly as they pertain to the control and functioning of the security agencies. Hence we may speak of regime rearrangement, rather than political transformation or democratisation.

Seen from this angle the process of demilitarisation seems hardly definitive or irreversible. And reversals and setbacks have occurred. The number of coups and military political actions that have occurred since 1993 suggest the ominous possibility that the pendulum may well be swinging back toward at least limited forms of remilitarisation. The movement away from the militarisation of politics in some African countries has coincided or overlapped with new manifestations of militarisation in others (or sometimes in the same countries), and been punctuated in situations such as Somalia, Liberia and Rwanda by yet more horrific instances of carnage and ethnocide far exceeding anything witnessed in the previous decade. On the other hand, there is encouraging evidence of democratic survival and even consolidation, with Benin (the pioneer in militarisation in the 1960s and 1970s and demilitarisation in the 1990s) once again leading the way with its second peaceful democratic succession in the 1990s (for good measure even returning power – through the ballot – to the very military dictator removed in the earlier democratic revolution!).

While this record of transition from military rule appears incomplete and even disappointing, it nevertheless needs to be placed in historical context to be evaluated properly. Before the recent round of democratisation, in only a few instances had military regimes in Africa dissolved their governments and actually handed over power (as opposed to civilianising or 'constitutionalising' it in some way), and these were invariably short-lived. Seen from this historical standpoint, the present processes of democratisation represent cautious evidence of progress. What is less readily apparent from the cases of transition cited above, is the depth of the political transformation which has taken place even in those situations where the military continues to cling to the shadow of power. This is manifest in the intensity of the popular pressures which have forced these political openings, the changes in the international and geopolitical context, the extent of the political liberalisation which these autocrats have had to endure (including real concessions to political pluralism and freedom of the press), and in the extent of the legitimisation crisis facing both military rule and military institutions.

How do we deepen and sustain this incipient process of demilitarisation of African politics?

Militarisation has reflected a daunting kaleidoscope of problems in Africa: economic mismanagement and decline, weak national integration, ethnic and class conflict, weak political institutionalisation, lack of political accountability, corruption, abuse of human rights, and so on. Name it, and it has been cited as a cause for one coup or other. Further, these 'causes' have been combined in shifting and highly specific combinations that can only be understood in their particular national and historical contexts. Militarisation is therefore a complex and highly dynamic phenomenon – a moving target – and therein lies both its analytical challenge and conjunctural political appeal. This has led to the suggestion that demilitarisation can only be sustained when these fundamental problems have been solved. This argument, however, is specious, not only because it potentially commits Africa to indefinite militarisation, but also because it fails to make a relevant distinction between, on the one hand, short term and long term tasks of democratic consolidation, and, on the other, between social and structural problems and political and institutional ones (the latter may be defined as those structures evolved for dealing with society's underlying problems and the conflicts arising from them). True enough that, as Magyar reminds us, "coups are merely the most frequent expression of utter frustrations encountered in nationbuilding in Africa, attempted under the most severe socio-political conditions." 3What is flawed, however, is the implied inevitability of coups. A rigorously 'structuralist' or reductive approach fails to explain why poor countries, such as Tanzania, were able to avoid military coups, while others more favourably placed on the socio-economic ladder succumbed. Socio-economic conditions are not in themselves a reliable or sufficient indicator of the likelihood or otherwise of coups. The case against military involvement in and domination of African politics is not only that it has failed to address problems at the structural level, but that it is by its very nature antagonistic to the development of the political and institutional framework required for peaceful resolution of conflicts and for mobilising national energies to tackle these problems. The argument can thus be reversed: without an end to militarisation, no solution to Africa's problems will be possible.

Over the long term, demilitari-sation and democratisation will be sustainable only if solutions to the underlying problems destabilising African politics and society are found. In the short to medium term, however, governance, institutionalisation, and coalition-building will be the crucial variables in this process. Among these institutional and political issues the most urgent, yet frequently neglected in Africa, is the one that bears most directly and immediately on the issue of demilitarisation: the question of the democratic subordination of armed forces and security agencies. It is on this issue that this article focuses.

The issues and the challenges posed by civil-military relations in the democratic transitions, the institutions and roles required to respond to them, and the political coalitions and strategies required to undergird them will be discussed below. The approach is influenced in part by Stepan's concept of a "democratic strategy toward the military."4 This strategy emphasises the negotiation of pacts between contending civilian interests, as well as between civil groups and the military, the need for civilians to empower themselves to challenge the claims of the military establishment to military secrecy and exclusive expertise over military affairs, and the demarcation of specific roles for civil society and political society in the democratic subordination of the military.

THE ISSUES

Democratising regimes and societies in Africa are confronted with a number of military-related issues of a constitutional, budgetary, and strategic character central to democratic consolidation. The first is the question of how to restore and sustain civil supremacy over the military following a period of military rule, and the appropriate institutional framework for this purpose. What roles should be played by the various political bodies, in particular the executive and the legislature? Second is the question of the military's future relationship with the political process. In what form and to what degree, if any, is the military to participate or be represented in the political process? What corporate political rights (if any) are to be retained by the military? Third is the issue of military privileges and prerogatives: the military's professional autonomy, salaries, wages and allowances, and the relative areas of application of military as opposed to civil law. And fourth is the problem of redefining the role and mission of the military, including its functions (if any) in relation to internal security. This will cover the areas of military doctrine, force levels, appropriate institutional structures and relationships, and equipment types and levels, and, where necessary (as in South Africa), conversion of defence industries and facilities. The objectives of reform should be at least four-fold:

the democratic subordination of the military and inculcation of respect for democratic values;


budgetary sustainability of the defence establishment;


improvements in military efficiency and capability; and


internal (institutional) stabilisation of the military.

Reform must thus integrate both operational and political objectives and concerns, not one to the exclusion of the other.

Newly democratising regimes in Africa face many challenges in attaining these objectives of subordinating unruly military establishments and restructuring dilapidated and inefficient military infrastructures. The reorganisation of civil-military relations is also likely to take place in the context of a multi-faceted crisis: economic crisis and resource constraints, constitutional disputes, human rights issues, contested legitimacies, and so on. These objectives call for a fundamental reassessment not only of structures of civil-military linkage and control, but also of force doctrine, mission and structures, even the existing military model as such.

How are these objectives to be achieved?

AFRICAN DEMOCRACIES MAY DRAW ON A VARIETY OF MODELS OF CIVIL-MILITARY RELATIONS

Several options and frameworks are available to democratising regimes in attempting to control their military forces. Welch5 identifies five of these devices:

the imposition of constitutional restraints on the role of the armed forces;



the delineation of clear spheres of responsibility between the civil and military spheres;



single-party penetration and cadre control of the armed forces;



ascriptive approaches designed to integrate the military into or separate it from society; and



deliberate minimisation of the size of the armed forces in order to curb its political power and influence.

The first two approaches conform to Huntington's6 concept of 'objective controls', while the others are consistent with his concept of 'subjective controls'. While Huntington sees these as the two main options available to political leaders in trying to control their armed forces, in real life such approaches have been further combined in various formulations. In addition, the professional 'model' comes in many forms. Victory in war presents further options for abolishing or disarming the military (Costa Rica, Haiti), or displacing or integrating the pre-existing military with a party militia (Zimbabwe). African democracies thus have several competing models of civil-military relations, indigenous as well as foreign, to look to. However although any of these may accomplish the twin goals of civil supremacy and political stability, not all are considered consistent with democracy.

The model traditionally regarded as most compatible with democracy is the so-called 'objective control' model – although it is suggested in the conclusion that such a model may be possible only in the long run, and that immediate tasks of democratic consolidation may require a more nuanced approach combining 'objective' and 'subjective' devices. The 'objective' or professional model incorporates institutions and professional norms that, on the one hand, inculcate acceptance of and respect for civil supremacy on the side of the military, and on that of civilians' respect for the professional autonomy and expertise of the military. At the same time this model provides institutions and channels to allow the military and security agencies to influence the political and decision-making process in a legitimate manner for professional ends. Reprofessionalisation implies the reorientation of the military from political involvement or internal security duties to an external defence posture and acceptance of civil supremacy. The price for this may be adequate (and sometimes much higher) levels of budgeting, training and equipping of the armed forces to allow it to properly satisfy this role. Reprofessionalisation and acceptance of civil supremacy have sometimes been facilitated by an exceptional or fortunate set of external circumstances. In the case of the extrication of the Greek army in 1974, it was the threat of war with Turkey (which forced the army to refocus its attention on traditional defence matters at the same time as its domestic political influence was collapsing disastrously), while in the case of Spain it was membership of the North Atlantic Treaty Organisation (NATO), which both brought it into an alliance emphasising professional norms and provided greatly expanded resources for modernisation.

In sum, democracy implies the political emasculation, but not the political impotence of the military. In withdrawing from formal power, the military exchanges overt political intervention for influence exercised through regularised and institutional channels. Indeed, it is arguable that the armed forces in the United States have more political influence and avenues for exercising it, than the military institution in any African military regime. For the military institution (as opposed to individual officers), this may actually enhance overall political influence. In certain countries, such as Sweden and the Netherlands, military officers and ranks are allowed to form unions and thus to develop their own channels of influence separate from the military command.7 Paradoxically, coups arise precisely because of such a lack of legitimate channels of communication, particularly for middle and junior level strata. On the other hand, however, many officers discover that overt intervention in politics carries a potentially high price: the erosion of esprit de corps, the development of cleavages among military strata, and – most galling of all – marginal influence over decision-making processes within the military regime.

THE NEED FOR EFFECTIVE INSTITUTIONS OF CIVIL COMMAND

There is an urgent need for African countries to think deeply and thoroughly about institutional arrangements for optimising regulation and control of the military. What should be the institutional channels for exercising civil command? How should powers be distributed between the military high command, the executive and the legislature? How should ministries of Defence be organised and what should be their function? What role should the civilian authorities play in determining such issues as defence policy and mission objectives or military procurement? How can legislative oversight be maximised? How can institutions in the new democracies best embody the twin principles of military expertise and civilian supremacy?

'civilian control' extends well beyond simply keeping the military in barracks. It implies (particularly though not exclusively on the side of the executive):

imposing effective oversight of military budgets, administration, procurement, and overall operations;



intervening effectively to shape, redefine or modernise the existing security doctrine, training programmes (including the content and curricular of service schools), force mission, goals and structures to make them consistent with democratic principles;



improving information and intelligence as the basis for effective monitoring and expertise in military and defence issues; and



playing an active role in the internal stabilisation of the military.

In short, civil authorities must take the lead in reshaping strategic concepts, relationships and ways of thinking. For this reason Virgilo Bertran has argued that the "education and training of a staff of qualified civilians in military planning and administration, logistics, etc. [should be] one of the first priorities [of any civilian government seeking the ability to] formulate, execute and control Defence and Military Policies."8

In most of Africa, institutions of civil command are weak and poorly equipped to execute the key roles that democratically elected authorities must play. In many African countries, legislative oversight of military and defence issues was one of the early casualties of independence. Ministries of Defence are typically staffed by junior level civil servants and clerks (the majority of whom are female) and have little strategic or supervisory role. In essence, the ministry is the armed forces, with a lonely minister sitting ceremonially atop of the structure. Many African constitutions still make formal provisions for such oversight in relation to the armed forces (or more appropriately, of budgetary matters pertaining to the armed forces), but are usually silent on the intelligence and other special security agencies that often form the underpinnings of authoritarianism and human rights abuses. Military and intelligence agencies may not only constitute a 'state within a state', they also maintain extensive and specialised contacts across international boundaries and constitute global networks parallel to and usually autonomous of those of the political and diplomatic authorities. As the Senate Intelligence Committee and the White House recently discovered with the CIA in Guatemala, these contacts and the 'messages' that they transmit may contradict those of the official authorities, but are almost impossible to monitor effectively. Because of their covert and extralegal character, the extraordinary reach of modern surveillance and intelligence gathering techniques, and their lack of public accountability, intelligence agencies present particular challenges to democratic freedoms, especially in their ability to circumvent the legal protections and restraints associated with democracy.

THE NEED TO DISTINGUISH BETWEEN CIVILIAN CONTROL AND DEMOCRATIC CONTROL

The division of labour within government and society in relation to oversight and control of military and security forces is a crucial issue in any democracy. civilian control of the armed forces in Africa has traditionally been exercised through the executive as opposed to legislative or judicial branches. In a democracy executive control is characterised by two key principles. The first is the use of force in a manner in keeping with the maintenance of constitutional and civil rights. The second is that in a democracy the exercise of key functions of oversight and regulation is carried out by or shared with the elected representatives of the people on the one hand, and the media and civil society on the other. Legislative self-empowerment with regard to military issues and policies is necessary for effective civil and democratic control, and includes the ability to effectively review defence objectives and military budgets and expenditures. This in turn, requires the development of the legislature's own research, information and monitoring expertise, and an appropriate committee system and congressional staff.9 The congressional system in the United States or the legislative system in the Netherlands,10 both of which allow elected representatives to extract information from the executive and military brass, hold their own hearings on military and intelligence issues, communicate directly with the armed forces, and otherwise develop expertise and oversight independently of the executive, should seriously commend themselves to African governments and legislators. For the military, on the other hand, the advantage is to broaden access to decision-makers and allies outside the executive sphere. Both of these two elements of the legitimate use of force and legitimate governance have been conspicuously lacking in authoritarian civil regimes. Hence, while civilian control is indeed an essential ingredient, it does not in itself amount to democracy as such.

THE ROLE OF THE CIVIL SECTORS

Another requirement for sustained demilitarisation is the existence of civil sectors, both organised and united enough on basic issues of principle to resist military intrusion on principle, while negotiating (where necessary) new relationships, principles, and guarantees with each other and the military. One of the main sources of the political power of the military regime is the ability to manipulate the cleavages in civil society. To forestall this possibility, Stepan suggests the revalidation of democracy pacts, with all democratic groupings committing themselves to the defence of democratic rights and processes, and desisting from inviting the military to intervene in political conflicts.11 A second function of the organisations of civil society (and this includes the media) is the need to develop autonomous military intelligence and the ability to challenge the claims of the military establishment to military secrecy and exclusive expertise over military affairs and, if necessary, to help restructure the prevailing security doctrine. As a rule, however, civilians in Africa have not been interested in security-related issues or (as is more often the case) have been intimidated from exploring them. In most African countries, institutions and agencies dealing with strategic studies (university centres, specialised research institutes, defence and intelligence organisations, and journals) are non-existent in the civil sector. On the other hand, Third World militaries tend to resist what they perceive as the encroachment by civilians on their areas of professional expertise (much of this difficulty should be removed by an appropriate redefinition of the concept of national security), yet they must be brought to accept that civilians have a legitimate and indeed essential role not only in overseeing military budgets and procurement, but also in helping to define the nature of security doctrine and overall defence policy. Hence defence and security-related issues are major areas of democratic struggle, and democratic movements must have a feasible strategy for forcing these issues onto the public agenda.

RESTRUCTURING THE MILITARY AND DEFENCE FRAMEWORK

The challenges facing democratisation in respect of the armed forces and security and intelligence agencies cannot stop at redefining the military's future relationship with the political process, but must also extend to a redefinition of the role, mission and doctrine of the armed forces. One of the key problems posed for democratisation is the weakened state of African armies. There is clearly a fundamental institutional crisis of the military in Africa, or perhaps to put it more accurately, a political, economic and military crisis of the inherited colonial model. One frequent reason for this is mismanagement. While the claim is often advanced (by scholars and military officers alike) that, by the very nature of their profession, armies and their officers are superior managers, in reality many African militaries are fairly atrociously managed, with outdated equipment, logistics and procurement systems, doctrines, and poor maintenance. Intra-military conflicts of an ethnic, generational, rank and ideological character, have often spilled over, in turn, to threaten the political order. The result is that African armies are poor producers of their primary commodity, 'national security'. On the contrary, they have become the primary source of national insecurity. At the same time, however, many of Africa's weak armies are involved in a large variety of internal security operations (against Islamic fundamentalists in Algeria, the Casamance rebels in Senegal, the Tuaregs in Mali, the RUF rebels in Sierra Leone, and so on). Several African armed forces have been locked in unwinnable wars for years; others have capitulated and disappeared altogether, while yet others have been reduced to just one of many compelling armed factions.

It is tempting to see this weakness of the African military as facilitating democratic transition, and to some extent this is true. However, this argument misses the paradoxical and contradictory nature of the relationship between military power and democracy. Few democracies have existed on the basis of military weakness.

Democracy cannot be expected to flourish where the state lacks the repressive and defensive capability required to effectively control its national population and at the same time defend its territory. National security in the true sense (not, as is often the case in Africa, regime security), and the assurance of law and order are legitimate and indeed essential aspects of democracy, which only states that are both strong and responsive can assure. For this reason there appears to be a strong correlation between democracy and strong national security capability (conversely, there are many examples where this capability – or its pursuit – directly threatens democracy and human rights). The paradox of democracy is that it demands a well-armed army that is at the same time subordinate to civilian control. Unfortunately, the corresponding African paradox is that of weak armed forces which are politically dominant.

Lack of professionalisation challenges democratisation in another way. The efficacy of civilian control depends in turn on the authority and control of military commanders over the institution. The concept of 'civilian control' becomes problematic where professional structures are weak or non-existent. Communication, and command and control systems in African militaries are typically weak and unstable. This poses dangers for both the civil government and the military hierarchy. Most coups proceed not from the high command (which is both the target and medium of civilian control), but from the middle and junior ranks of the military (the relatively faceless subaltern levels) and from the activities of factions, some of which are able to establish themselves sufficiently to take power on more than one occasion. Ironically, the military tends to be particularly vulnerable to this kind of factional blackmail, both because of the overwhelming emphasis on 'unity' mandated by its professional ideology and the reluctance of soldiers to take up arms against their own comrades.

The issue of the professionalism of the military is also crucial in democratic transitions, because the more professional the military institution, the more likely it is to respond to political change with a coherent strategy to protect its core professional interests (vide the South African military and military-industrial complex). On the other hand, the less professional the military, the less likely it is to distance its interests from those of the regime or social and political groups (Togo, Zaire and Burundi).

MILITARY BUDGETS AND EXPENDITURES

Budgetary questions are a key issue in democratic transitions, since they are at the core of both military privileges and military reform. In general, real military outlays declined substantially in most African countries throughout the 1980s and early 1990s. The problem in the African context, however, is that while real military spending in many countries has stagnated or declined to the point where many armies are barely able to function, it is nevertheless beyond the range of affordability. Resource constraints are going to be a critical issue in this area. In many countries it is not uncommon for soldiers not to be paid for months on end; since 1992 there have been pay revolts in Niger, Zaire, Lesotho, Mozambique, and most recently Guinea and the Central African Republic. Nevertheless, 'Faustian bargains' with the military may be required to stabilise democratic rule, and this may well include increased defence expenditures, which may in any case be required to underwrite the needed reforms. In itself, however, budgetary guarantees will not necessarily be sufficient to deter coups. On the other hand, drastically reduced budgets immeasurably complicate the problem of reform, unless they are combined with some form of downsizing and thorough reform of military structures. Budgetary policy thus needs to be co-ordinated with military restructuring designed to reduce the size and manpower of armies, while at the same time improving their armaments, logistics and service conditions.

REDEFINING NATIONAL SECURITY CONCEPTS AND STRUCTURES

There is a need to redefine the concept of 'security' so that it ceases to be the exclusive preserve of the military, by, for instance, drawing a distinction between military and non-military aspects of security or between 'security' (which is the responsibility of the citizenry as a whole) and 'defence' (the legitimate area of the armed forces). In general, existing security doctrines need to be reformulated to eliminate their

preoccupation (even in the absence of external threat) with arms procurement and the emphasis on purely military considerations – thus privileging securocrats – and more on non-military (social and economic) aspects of security (the satisfaction of basic human needs for food, shelter, education and other social services);



the corresponding dependence on foreign sources and strategic concepts and relationships;



their class, ethnic and gender biases (viewing men as the heroic defenders and women as the victims), and



to minimise the corrosive competition between regime security (which frequently leads to the deliberate emasculation of the defence posture) and the security of the nation as such.

Some of these considerations are at the basis of the vibrant debates on security currently taking place in South Africa.

Throughout Africa attitudes on military questions remain fluid, though critical. Important questions are being posed in the light of the crisis of existing military models. Is a military organisation rooted in civil society, rather than isolated from it, the answer? Would the military be more accountable if it were less distant from society? Should the military realistically be integrated into the political process in some way to stop its irregular interventions (an argument that finds some support from pre-colonial African structures)? Can the military be a force for democratic governance if it is not itself internally democratic? Certainly, few Africans are wedded to the existing post-colonial military format, particularly given its economic and political exactions. Africa has seen all the costs and few of the benefits associated with modern standing armies. However, the nature of military discourse suggests that these are issues also being debated within the military itself. Military orthodoxy on some of these questions has been challenged by sections of the military, particularly among the subaltern ranks. Military reform may also possibly present an opportunity to enhance internal democracy at the same time as it improves military efficiency and institutional stability.

STRATEGY

Intervention by civic organisations in the area of civil-military relations not only requires clearly thought-out positions but also feasible strategies to carry forward these solutions, including the formation of the necessary political coalitions to support this process. Interventions by democratic movements in the area of defence must involve a carefully calculated strategy (a separate issue in itself), with similar sensitivity to questions of timing and execution. Information and intelligence have already been mentioned. To cultivate this effectively, the democratic forces must (as Stepan again reminds us) have some insight into the nature of current discourses within the military. Contact with sympathetic retired military officers can give some ready insight into this. Second is the need to construct appropriate coalitions, which must as far as possible include allies from the military favouring return to professionalism. The basis for this must be the consensus that the dictatorship has been disastrous, not only for democracy, but also for the military institution. This insight suggests a key opening around which the new democracies can fashion a strategy for subordinating the military, by appealing to the professional self-interest of the military. By the same token it can be demonstrated that democracy, through its insistence on civilian subordination and political non-involvement of the military, may well form the best foundation for military professionalism and specialisation. There should also be the recognition that the military can continue to resist democratisation only at the expense of its own professionalism and ultimate survival (Haiti is a case in point). Forging a clear consensus between the military and civilians on such issues as force levels and mission goals and on the overall direction of reform, is essential to give broad legitimacy to the reforms. Any negotiations or pacts with the military must, as far as possible, be subscribed to by all democratic organisations. Although considered inherently anti-democratic, 'pacts' have often been seen as one way of reconciling the divergent interests of civilian groups, factions and parties on the one hand and that of the military on the other. However pacts on this scale have been relatively rare in the African transitions. South Africa and Zimbabwe were important exceptions. This is probably due to several reasons, among them the absence of coherent and professional military forces; the absence of strong and indispensable socio-economic élite groups; lack of experience of the military in working with such civil élite groups, and overall its lack of structural rootedness.

ENDING THE ISOLATION OF THE MILITARY

Finally, there is a need for civil society to construct its own dialogue supportive as well as critical) with the military. The rift between the military and civil society is especially large in Africa. Biddle, arguing for 'interaction and communication' between the military and the civil sectors, suggests that "[t]he military must not be isolated and insulated from civil society. Rather the two must have an active and ongoing exchange. If the military is exposed to the diverse interests and forces within pluralistic civil society, the armed forces begin to understand that their role is not to dominate or influence the civil sector but to be subordinate to it and to protect it from foreign threat. If the military learns to appreciate the broad spectrum of interests that exist, it intuitively realizes that its function is to remain neutral and thus serve democracy. Therefore, civil society and the armed forces must be constantly exposed to one another and maintain an open and transparent dialogue."12

The lack of linkages between the military and civil sectors and national economy deprives the military both of the ability to develop a rational calculus toward the domestic political and social order, and at the same time the means to diffuse influence through broad-based linkages and alliances with the civil sector. Where these manifold connections exist (through a military-industrial or intelligence-industrial complex), the military becomes embedded in civil life; military spending has the effect of fusing or integrating the military and civil sectors. Where they do not, the military is isolated from social and economic currents (a situation which favours persistent coup activity).

Retired military personnel can play a key role in mediating between the military and civil society. In many African countries retired military personnel, most still in their prime, outnumber those in active service, and play important roles in civil and economic society. While large numbers of retired or demobilised military personnel must raise security concerns, and governments have tended to be rather sensitive to their activities, they do form a source of expert knowledge and intelligence on military and defence issues which can be made available to political leaders and civil society. They are strategically located in being the only ones with exposure to and comprehension of both civil and military life and thus in a unique position to mediate between the two sectors. Retired military personnel may consider forming lobbies and public interest associations, collaborating in defence research institutes, and contributing to legislative hearings and public forums.

REGIONAL SECURITY ARRANGEMENTS

With Africa's small nations, weak and small armies and limited defence budgets, regional defence arrangements and mutual defence pacts seem essential for rationalising defence costs. However, progress in this area has been slow to materialise in part because of the OAU's principle of non-interference in the affairs of member states. Hence, although African armies have long been involved in international peacekeeping, it is only now that they are becoming involved in regional peacekeeping (and even more recently are they doing so under their own steam). In the absence of similar self-imposed constraints most armed interventions have been carried out or co-ordinated by foreign powers or the United Nations, even when they involved African troops. However there is evidence that this reticence is breaking down, and that African states, whether unilaterally or collectively, are looking more seriously at the possibility of intervention across neighbouring borders in extreme situations. The ECOMOG intervention into Liberia is a landmark in this respect. A particularly significant dimension of this development may be an emerging willingness to use collective intervention to frustrate illegal coups (and protect democracy) in neighbouring countries, such as the reversal of the 1994 attempted coup in Lesotho, the pressures on the military regime in Niger to restore power to civilians after the January 1996 coup, and the current threat of sanctions against Burundi by regional trade partners. Because of the long collaboration between the front-line states and the historical absence of coups, Southern Africa provides the best opportunity for this kind of co-ordinated action. On the other hand, the events cited here have also shown that the necessary consensus and co-ordination cannot always be expected to exist either among regional partners, or between regional and international ones (the same Western nations that leaned heavily on the regime in Niger to hand over power have been much more lukewarm about supporting the sanctions against Burundi). Regional integration of military forces may also make it more difficult to execute a coup against a national government.

ROLE OF EXTERNAL FORCES

External forces can and have played a crucial role in military restructuring, particularly in brokering peace agreements and providing military retraining and financial and technical aid to accelerate the resettlement of demobilised military personnel. On a non-military level several agencies, such as USAID and the National Democratic Institute for International Affairs, are helping to promote the development of national policy and policy dialogue, in addition to addressing institutional issues. Less formally, other European governments, such as those of Scandinavia, are evolving programmes that may focus on NGO action to help sustain demilitarisation and institutionalise stable civil-military relations. In one of the earlier instances of such external support, the National Democratic Institute helped to sponsor an international conference in Montevideo which brought together military and civilian experts from Argentina, Uruguay, Brazil and other countries to address issues of defence and control of the military in a democratic society, ending in the issue of the Montevideo Declaration of July 1989, with its valuable guidelines on the role of the legislature.13 Recently, both the International Mometary Fund and the World Bank have openly begun to address issues of military spending and reform of defence budgets.14

PUNISHING THE OFFENDERS?: DEALING WITH HUMAN RIGHTS ABUSES

A key issue in sustaining demilitarisation is how to deal with political crimes. How should new democracies deal with atrocities and human rights abuses by government and security agents, or with those who overthrow constitutional governments through force of arms? Should the emphasis be placed on justice and accountability (if not retribution), or on national reconciliation? This question has been posed in several of the current transitions – the Truth Commission in South Africa, the trials of the former Dergue in Ethiopia, of Traore in Mali and Hastings Banda in Malawi, and the attempted prosecution of the perpetrators of ethnocide in Rwanda. As is already known from Latin America, this issue can weigh heavily on the transition and colour the prospects for civil-military relations. As with other issues, there are few routine or comfortable answers to such questions, and many compelling arguments can and have been made on both sides. Uganda has inserted a clause in its new constitution to the effect that anyone illegally overthrowing a constitutional government should and will be punished, no matter how far down the road. Until recently this would have seemed like quixotic self-indulgence. However, recent events in South Korea will probably change the terms of this debate. By convicting Chun and Roh for their part in the Kwangju massacres and imposing severe sentences on them, Korea has sent a brilliant and unambiguous reminder – the more so since both culprits had been highly successful in leading the modern economic and political transformation of Korea – that history can always be revisited. If Chun and Roh, why not Mobutu and Eyadema?

CONCLUSION: WHAT FORMS OF CONTROLS AND CIVIL-MILITARY RELATIONS ARE MOST APPROPRIATE FOR AFRICAN POLITICAL SYSTEMS?

In the final analysis, however, it must be remembered that it is not laws or constitutions, but realpolitik that underpins successful civilian control of the military in political transitions. Some African leaders have successfully combined a variety of stratagems – co-optation, bribery, political penetration, ideological indoctrination, armed counterweights, ethniticisation of the security services, and other divide-and-rule tactics - to subordinate their militaries. The lessons from these histories, while sometimes unsavoury, may nevertheless be pressed in the service of democratic consolidation.15 It is simply impossible to determine on abstract grounds what form of civil-military relations is best for all African countries. Each national case is distinctive, though certain democratic principles will have to be kept clearly in mind. Good governance is clearly the key variable, but not even the most committed leadership can forget the lessons of realpolitik in the presence of arms. In any case it is doubtful that, at least in the early stages of democratic transition, formal constitutional structures alone can respond to the imperative of preventing coups and re-establishing control over military forces. Other devices may need to be put in place pending the development of normatively-oriented constraints.16 This is an inherently dangerous strategy, but such 'subjective' forms of control seem to have worked relatively well when confronting armies with weak professional traditions. Some of these strategies, such as armed counterweights, are not always inconsistent with constitutional control, although they are more feasible in some security formats than others, namely in the French military model with its Republican Guards, gendarmerie, customs police, etc., in addition to the regular military, as opposed to the British military model which lacks a similar tradition of multiple armed bodies.

ENDNOTES

This paper is adapted from several previous papers by the same author as a basis for discussion at the Africa Leadership Forum conference on The Military and Civil Society in Africa, Lilongwe, Malawi, 23-25 September 1996. It draws extensively on material in E Hutchful, Militarism and Problems of Democratic Transition, in M Ottaway, Democracy in Africa: The Hard Road Ahead, Lynne Reinner, Boulder, 1997 (forthcoming); E Hutchful, Military Issues in the Democratic Transitions in Africa, in E Hutchful and A Bathily (eds.), The Military and Militarism in Africa, Codesria, Dakar, 1996 (forthcoming); and on the results of work done as a contributor to the project on Comparative Civil-Military Relations: Understanding the Mechanisms of Civilian Command in Small Democracies, of the Arias Foundation.



Burundi's 'creeping coup' ended with the return and takeover of the government by former Major Pierre Buyoya in July 1996.



K P Magyar, Military Intervention and Withdrawal in Africa: Problems and Perspectives, in C Danopoulos (ed.), From Military to Civilian Rule, Routledge, New York, 1992.



A Stepan, Rethinking Military Politics: Brazil and the Southern Cone, Princeton University Press, Princeton, 1988.



C Welch (ed.), Civilian Control of the Military: Theory and Cases from Developing Countries, State University of New York Press, Albany, 1976.



S Huntington, The Soldier and the State: The Theory and Politics of Civil-Military Relations, Random House, New York, 1957.



In spite of what may be thought to be their 'revolutionary' implications, serious consideration should be given to extending similar trade union and collective bargaining rights to the African military and particularly to the ranks. The notion that potentially the most politically powerful social sector must be denied the most elementary economic rights, including the right to articulate its economic needs, is a quite outmoded one (which, among other factors, dangerously reduces the intelligence available to the civil authorities).



V Bertran, Some Lessons Learned in Civil-military Relations in Argentina, paper read at Conference on Civil-military Relations in Latin America, the Democracy Projects, School of International Service, The American University, 4-6 May 1995.



Stepan, op. cit.



G C de Nooy, Comparative Civil-Military Relations in Small Democracies: The Case of the Netherlands, paper read at the Seminar on Comparative Civil-Military Relations: Understanding the Mechanisms of Civilian Control in Small Democracies, the Arias Foundation for Peace and Human Progress, San Jose, Costa Rica, 1-3 August 1996.



Stepan, op. cit.



G C Biddle, A Principle Lesson of Civil Military Relations, paper read at Conference on Civil-military Relations in Latin America, op. cit.



National Democratic Institute for International Affairs, Towards a New Relationship: The Role of the Military in a Democratic Government, National Democratic Institute for International Affairs, Washington, 1990.



See for instance, G Lamb and V Kallab, Military Expenditure and Economic Development: A Symposium on Research Issues, World Bank Discussion, 185 and IMF Survey, 14 December 1992. Among the research projects of the World Bank is one dealing with the economic implications of the transition from war to peace and another with the best ways to downsize armies.



At the same time, such structures, which have protected African autocrats from coups, have tended to fall victim to democratic transitions. An example of this was the system of party control over the Tanzania Peoples Defence Force, which effectively preserved civil supremacy after the 1964 mutinies, only to be dissolved with the introduction of political pluralism. The apparent irony is that, in order to facilitate democracy, the established and highly successful structure of civilian control – a crucial ingredient of democracy – was discarded.



J Frazer, Conceptualizing Civil-Military Relations During Democratic Transition, Africa Today, 42(1-2), 1995.