1. Introduction: Democracy and the Military in comparative perspective

R.J. May

Stephanie Lawson

Viberto Selochan


The authors aim, within the context of current discussions of ‘transitions to democracy’, to examine the central concerns of the subject. They approach the topic within the framework of a larger interest in the process of regime change and regime maintenance in Asia and the Pacific since it is clear that the military has played a major role both in bringing about changes of regime and in forestalling change.

The principle questions addressed are first, what role has the military played in regime change and maintenance in the countries of Asia and the Pacific, and second, have differences in the degree of military involvement in politics been systematically associated with differences in the performance of the political system, particularly its performance in relation to democratic criteria?

Their sense is that the military is likely to continue to play an important role in the politics of the countries of Asia and the Pacific, notwithstanding tendencies towards democratisation. They propose a shift in focus of research from the military per se, to the activities of soldiers in the complex of military-civil relations.

From the processes of decolonisation which dominated the political history of Africa, Asia and the island Pacific in the mid twentieth century, most post-colonial states emerged with constitutional structures inherited from, or at least heavily influenced by, the Western democratic models of former colonial powers. Among the principal general features of such constitutions were: separation of the legislature, executive and judiciary; popularly-elected legislatures in which competitive political party systems were expected to provide the basis for a division between government and opposition; and the subservience of the military (whose primary role was generally seen to lie in defending the country against external aggression) to the civil authorities.

In the early stages of decolonisation it was expected that indigenous armies, following the models set by the metropolitan powers which created them, would refrain from direct involvement in politics. Nevertheless, even in those newly independent states in which the military did not gain a political inheritance by virtue of its role in the winning of independence, rather than imbuing the armed forces with a military professionalism which required absolute obedience to the civil authority, colonial rule left behind armed forces more often oriented towards maintaining internal order than to external defence, and therefore implicitly attuned to domestic politics. This was particularly evident in states marked by strong ethnic cleavages, where colonial policies often involved the recruitment of military personnel from those ethnic groups which appeared most compliant (see below).

In fact, shifts from parliamentary democracy to one-party or military dominated regimes were not long in coming. Africa had its first military coup in 1958 and there were coups in Burma, Thailand and Pakistan in the same year. A torrent of military interventions followed during the 1960s and 1970s. Between 1945 and 1976, Nordlinger (1977:xi) estimated, more than two thirds of the countries of Latin America, Asia, Africa and the Middle East had experienced varying levels of military intervention. A study of sub-Saharan Africa between 1960 and 1982 alone recorded 90 plots to overthrow governments, 60 attempted coups, and 50 successful coups (Orkand Corporation quoted in Seitz 1991:65). In 1977 La Palombara commented: ‘Military coups are now so frequent and widespread they must be considered as significant as elections’ (‘Foreword’ in Nordlinger 1977:x); even earlier, Janowitz (1971:306) wrote:

The intervention of the military in the domestic politics [of non-Western states] is the norm; persistent patterns of civil supremacy are the deviant cases that require special exploration.

Because military interventions were widely seen as a denial of the democratic values and institutions which retiring colonial powers had hoped to establish in the new states, considerable scholarly attention was devoted to explaining why and how military coups occurred. Explanation was sought in the motives of coup leaders, the structure of the military, and in predisposing and facilitating socio-economic, political, and external conditions.[1]

Early scholarship sought the reason for military intervention in the relative ‘underdevelopment’ of civil political institutions. More specifically, some writers argued that in new states the military typically was more cohesive, better organised, more ‘rational’, and more strongly committed to modernisation than the rest of society, including politicians, and that military intervention was a predictable response to the inefficient and often corrupt administration, and political fractiousness, which characterised the civil government in many new states.[2] For those who saw a strong state as a necessary precondition for economic develop-ment, military intervention was not necessarily a bad thing (for example, see Lefever 1970). Such a viewpoint, however, raised some big questions: in particular, if the military intervened because the institutions of civil government were ‘underdeveloped’ or not working well, what chance was there of civil institutions ever developing? Although military coup leaders frequently presented themselves as intervening temporarily, once out of the barracks they were seldom in a hurry to return; moreover the actions of military rulers – banning political activity, suspending constitutions, imposing media censorship, and so on – were frequently inimical to the development of civil politics.

An alternative line of explanation saw military establishments as motivated less by a culture of rationality, sound management, and modernity than by its corporate interests. Military intervention was especially likely, they argued, when the military was marginalised or fiscally deprived, or its interests, autonomy, or ‘professionalism’ threatened. (See, for example, Janowitz 1964; First 1970; Bienen 1971; Hakes 1973; Thompson 1973; Nordlinger 1977; Horowitz 1980; Rouquié 1987.)

In both these approaches the military was seen essentially as a cohesive entity with a sense of collective identity. A third school of thought, in contrast, portrayed the military as simply an extension of the larger civil society, subject to the same class, regional and ethnic cleavages, prone to internal friction, and likely to side with particular political factions at particular times. Taking this argument further, Decalo (1976) suggested that the reasons for military coups were to be found in the personal ambitions of coup leaders. The idea that the military was at least potentially fragmented had particular salience in those states in which the military had a specific ethnic bias, often the result of deliberate colonial policies of recruiting from ‘martial races’ or from ethnic minorities rather than dominant ethnic groups which might thus be given the means to challenge colonial rule (Daalder 1969; Guyot 1974; Kabwegyere 1974; Mazrui 1976; Hansen 1977; Nordlinger 1977; Enloe 1980; Horowitz 1985; also see Gow 1991). The role of social class, on the other hand, was contested: while some saw the military as likely to pursue the interests of the middle class, others saw it as characteristically cutting across class interests. (Major contributors to this debate include Huntington 1968; Lloyd 1973; Halpern 1963; Nordlinger 1977; Alavi 1979; Luckham 1979; Perlmutter 1981; Nun 1967, 1986.) Inter-generational tensions, and rivalries between age cohorts and political factions within the military were seen to be increasingly significant as the number of coups – especially ‘second round’ coups – increased; Seitz (1991:70) estimated that ‘intra-military elite factionalism’ accounted for about a third of the plots, attempted coups and coups recorded in the Orkand Corporation study (see above).

Several studies distinguished various types of coup and coup attempt, ranging from those (typically first coups) which sought to set up new regimes, through internal military putschs, to ‘coups’ directed against regime change (for example, see Huntington 1968; Hoadley 1975; Chazan et al. 1988; Luckham 1991).

Of course, these various ‘explanations’ were not necessarily mutually exclusive: a state in which there was an imbalance in development between the institutions of state control and those of popular participation, for example, was probably more vulnerable to intervention to assert the military’s corporate interests. ‘Isolating “The Cause” of a coup d’etat’, Welch (1974a:135) suggested, ‘is a fruitless exercise. Personal, organisational and societal factors are intermingled’. Moreover, as Horowitz (1980:8) suggested, different explanations were sometimes appropriate to different levels of explanation (if in fact, they explained anything at all). Not surprisingly, then, a growing body of case studies provided support, in varying degrees, for all of these hypotheses, suggesting that while there were some recurring characteristics of military intervention, the explanation of individual cases required an understanding of their particular historical and social circumstances.

With military or civil-military regimes becoming increasingly the norm in Africa, Asia, Latin America and the Middle East, from around the mid 1970s students of the military began to shift the focus of their enquiry from explaining coups to a second enterprise, that of assessing the relative performance of military regimes. Early writings on military intervention in politics tended, as we have seen, to regard military intervention as essentially anti-democratic, but to see military regimes as probably more capable than democratic civilian regimes of achieving modernisation and development. A series of studies in the 1970s and early 1980s (for example, Nordlinger 1970, 1977; Schmitter 1971; Hoadley 1975; McKinlay and Cohan 1975, 1976; Jackman 1976; Zuk and Thompson 1982) addressed this question in fairly broad terms but found that, in terms of performance (variously defined), military regimes did not form a distinctive regime type. Heeger (1977:247) went further, suggesting that for Africa and Asia in the decade 1965-1975, ‘most military regimes have hindered the development of their countries’. More recently Seitz (in Kennedy and Louscher 1991) has concluded from a study of 38 sub-Saharan African states that there is ‘no significant discernible pattern separating the economic performance of military and civilian regimes’ (ibid.:7, italics added). Crouch (1985, 1988), addressing the record of the military and development in Southeast Asia for the period 1970-1985, also dismissed the particular role of the military as a decisive factor; he went on to emphasise the significance for economic development of maintaining political stability but concluded that in this respect, too, the military’s record was mixed.

Measures of political performance, on the other hand, seem to show a more definite pattern: Nordlinger (1977), for example, looking at four measures of political performance (legitimisation, noncoercive rule, minimisation of violence, and responsiveness to popular wishes), concluded that the performance of military governments ‘is significantly and almost consistently poorer than that of civilian governments’ (ibid.:197). More recently, Finer (1991), using Freedom House data, notes that all but two out of 36 military governments (i.e. 94 per cent) were ranked as authoritarian and lacking basic civil freedoms, compared to 60 per cent of 73 civilian regimes. Nevertheless, the only safe – if unexciting – generalisation seems to be that, as stated by Luckham (1991:22), ‘Military regimes are usually but not invariably authoritarian, and authoritarianism frequently but not always involves rule by soldiers’.

As more and more states came to experience periods of military rule it also became obvious that stereotypical models of military rule were inadequate. In some countries the military, or factions within the military, had simply made a blatant grab for power; in others the military intervened to replace an ineffective or corrupt civilian government with the stated intention of handing power back to civilian rule; in still others the military and civilian authorities established a system of joint participation in government. Consequently, a third major endeavour of the literature on the military in politics has been to differentiate types of military and civil-military regime. Janowitz (1964) made an early distinction between five types of civil-military relations, which he labelled authoritarian-personal control, authoritarian-mass party, democratic competitive and semi-competitive systems, civil-military coalition, and military oligarchy. Welch (1974a) suggested a distinction between personalist, corporatist and interventionary professionalisationist military regimes. Nordlinger (1977) distinguished military regimes by their role, as moderators, guardians or rulers. (Similarly see Perlmutter’s [1981] classification of arbitrator and ruler praetorian regimes.) Perlmutter (1980), arguing that, ‘The modern military regime is distinctly and analytically a new phenomenon, restricted to the developing and modernising world’ (p.96), suggested a fivefold typology, dividing military regimes into corporative, market-bureaucratic, socialist-oligarchic, army-party and tyrannical. Finer (1991), confining himself to countries in which the current regime is the outcome of a previous illegal usurption and in which the head of state is a member of the military, and adopting a more structurally-oriented classification, divides military governments into three sub-types: the military junta (or stratocracy), the presidential type, and those (perhaps more properly regarded as authoritarian civilian states) which, while founded by a military coup, have a civilian cabinet and a (limited) competitive party system and legislature.[3] What is emphasised by these (and other) authors, however, is not simply the variety of military regime types (or in Finer’s terms, subtypes) but the lack of a clear dividing line between military and civilian regimes. As Heeger (1977:243) put it:

It has become increasingly apparent that the rigid dichotomy between ‘civilian’ and ‘military’ regimes cannot be maintained … the transition from military rule can be seen in one sense as a transition from one mixed system to another mixed system.

Similarly, Finer (1982:282) argued that ‘the class of “military regimes” embraces a number of distinct subtypes which merge, gradually, into civilian regimes’, and Bebler (1990) proposed a continuum of civil-military relations, whose opposing extremes he called ‘civilocracy’ and ‘militocracy’ and whose middle ground was occupied by equal partnership arrangements, dual hierarchies, and ‘fused’ systems. Bebler went on to observe:

Whether officially recognised or not, the military everywhere constitutes an important part of the state apparatus and of the political system, and the soldiers, even when sound asleep in their barracks, participate in the political process and tacitly share political power with civilian rulers (ibid.:262-63).[4]

A further aspect of the discussion of military regime types lay in the recognition that the role of the military may change over time. Huntington (1968:221) observed:

As society changes, so does the role of the military. In the world of oligarchy the soldier is a radical; in the middle-class world he is a participant and arbiter; as the mass society looms on the horizon he becomes the conservative guardian of the existing order.

On the other hand, Welch and others suggested that once in power military regimes changed systemically – in Welch’s (1974a) analysis from personalism to corporatism to interventionary professionalism; in Perlmutter’s (1981) analysis, from arbitrator to ruler and back to arbitrator. Studies of the military in Latin America in 1970s suggested that a more fundamental, secular change was taking place in the military’s perception of its role: increasingly, Stepan and others argued, soldiers were taking on civilian roles of administration, management and economic enterprise.[5] Stepan (1973, 1978) referred to this as the ‘new professionalism’. Such a military role expansion was evident in Southeast Asia in the 1960s, and Lissak (1976:13), writing about Thailand and Burma, spoke of ‘the penetration of the officer corps, either collectively or as individuals, into various institutional fields, such as economic enterprises, education and training of civilian manpower, fulfilling civilian administrative functions, and engaging in different forms of power politics’. In Indonesia, the ‘civilianising’ of the armed forces had been anticipated even before 1960s.[6]

In part, the role expansion of the military in the Third World has reflected a shift in predominant concern, from external defence to internal security (embracing civic action programs and the growth of paramilitary forces).[7] But in part also it has been a strategy by which military regimes have sought to consolidate and legitimate their role in government, especially where that role has been challenged by civilians or external actors, or threatened by factionalism from within.

This suggested a further issue for investigation: the question of ‘exit’ – how can the army, once in power, be returned to the barracks? As early as 1962 Finer observed that, ‘In most cases, the military that have intervened in politics are in a dilemma: … they cannot withdraw from rulership nor can they fully legitimise it’ (1962:243). In fact, of course, some coup-makers did withdraw; indeed Finer (1985) later acknowledged that, ‘Most military regimes … have very short lives’, and went on to review the practice and theory of military withdrawal in terms of two principal alternatives – institutionalisation (essentially what other writers have termed ‘civilianisation’) and abdication. Following Sundhaussen (1984, 1985), Finer suggested that the conditions for military withdrawal parallelled, in reverse, those for military intervention, and identified two sets of dispositions and societal conditions for withdrawal; successful abdication, he concluded, required that the personal, corporate and ideological interests of the military be protected, and that the party or party system to which the military handed over be ‘organised, not unwise, and in effective control of the country’ (Finer 1985:30). Contemporaneously with Finer’s analysis of ‘the retreat to the barracks’, Clapham and Philip (1985) rephrased the dilemma for military regimes as being to develop a mechanism for succession without jeopardising their own supreme position; they saw six likely alternative outcomes – handback, civilian renewal, authoritarian clientelism, factional clientelism, and military party state, and ‘just another impasse’ (as when the military, under pressure, hands power back to a weak civilian state). (Also see Finer 1962; Huntington 1968; Welch 1971, 1974b; Bienen and Morell 1974; Heeger 1977; Nordlinger 1977; Needler 1980; Horowitz 1980; Third World Quarterly 7(1) 1985; Danopoulos 1988.)

However, as Heeger (1977:244) warned:

… in speaking of the military’s withdrawal from politics one risks exaggeration. The transfer of formal political power to civilians may be accompanied by a full-scale return to the barracks on the part of the military. More likely, however, is the emergence of the military in a somewhat less prominent, but no less political, role.

Typically, military personnel, having seized power, sought either to consolidate their position, penetrating civil society (sometimes setting up military-backed parties) and discouraging opposition, or to shift from a ‘caretaker’ role by restoring civilian governments while maintaining a guardian or veto role and strengthening linkages with civilian politicians and business people. Cases of a single military intervention, followed by consolidation or withdrawal, have in fact been unusual; more common have been cycles of greater and lesser military involvement of politics.[8] ‘Proclaimed intentions’, Finer (1985:17) observes,

… usually bear little relationship to the outcome. Rulers who intend to hand power back to civilians and do so are rare … Rulers who say they so intend but in fact hang on to power are more common … Rulers who make no promises to hand back, or openly propose permanent military rule are very common … But rulers of this intention who actually succeed in carrying it out are most uncommon.

At this point the literature on the military in politics converges with the burgeoning body of writing on regime change (see, for example, Linz and Stepan 1978; O’Donnell, Schmitter and Whitehead 1986; Diamond, Linz and Lipset 1988, 1990; Goodman, Mendelson and Rial 1990). Specifically, the recent perceived trend towards democratisation in parts of Latin America, Africa and Asia has revived interest in questions of military withdrawal, though as Luckham (1991:12) reminds us, ‘the installation of a military government [and, per contra, the withdrawal of the military from government] by no means always adds up to a change of regime’.

The questions raised here, and others, have, of course, been substantially addressed both at the theoretical level and in a growing volume of case studies, including comparative Asian and Pacific studies (among the latter, see Guyot and Willner 1970; Hoadley 1975; Zakaria and Crouch 1985; Olsen and Jurika 1986; Soedjati and Yong 1988; Heinz, Pfennig and King 1990; Selochan 1991b). In the light of the current discussion of ‘transitions to democracy’, however, and especially in view of the recent experience of some Asia-Pacific countries in resisting democratisation (Burma, China, arguably Indonesia, Singapore and Tonga) or moving away from it (Fiji, and arguably Malaysia), it seems worth revisiting some of the central concerns of the literature. More specifically, we have approached the topic within the framework of a larger interest in the processes of regime change and regime maintenance in Asia and the Pacific (see May 1994), since it is clear that the military has played a major role both in bringing about changes of regime and in forestalling change.[9]

The principal questions which this volume addresses, therefore, are, first, what role has the military played in regime change and maintenance in the countries of Asia and the Pacific, and, second, have differences in the degree of military involvement in politics been systematically associated with differences in the performance of the political system, particularly its performance in relation to democratic criteria?

Before turning to the case studies presented in this volume, however, it is necessary to reflect briefly on some key concepts.

[1] There have been numerous attempts to review the copious literature on military coups (see, for example Lowenthal 1974; Hoadley 1975; Nordlinger 1977; Perlmutter 1980; Ball 1981; Valenzuela 1985; Kennedy and Louscher 1991). We will not repeat that exercise here, though some features of the debate will be highlighted.

[2] Among a number of studies which broadly pursued this theme, major contributions included Shils (1962); Pye (1962, 1966); Finer (1962); Johnson (1962); Halpern (1963); Riggs (1964); Janowitz (1964); von der Mehden (1964); Huntington (1968); Zolberg (1968); Daalder (1969); Dowse (1969); Lefever (1970); Bienen (1971, 1983); Lissak (1976); Perlmutter (1977, 1981); more recently see Crouch (1985) and Chazan et al. (1988).

For some dissenting views see Lee (1969); Welch (1974a); Mazrui (1976). Mazrui in particular saw the military, in Africa, as likely to ‘retraditionalise’; similarly see Crouch (1979) on the ‘neo patrimonialism’ of the military in Indonesia.

[3] In an earlier paper, Finer (1982) presented a ‘morphology of [32] military regimes’, ranging from ‘military-supportive civilian regimes’, through ‘indirect-military regimes’, to ‘military regimes proper’, based on an analysis of ‘who governs’. Also see Luckham (1971) and Bebler (1990).

[4] Also see Finer (1962, 1985); Lee (1969); Lloyd (1973); Bienen and Morell (1974); Heeger (1977); and Perlmutter (1981); however cf. Luckham (1991: 2): ‘The more one looks at [the military], the more it decomposes like the vanishing smile of the Cheshire cat, into the turbulent social and political forces that swirl around it. Yet the more one seeks to explain its role in relation to those forces, the more its military specificity is brought (like the smile) back into focus’. Even Bebler, having introduced the idea of a civilian-military continuum, argues against those who deny the perceptual validity of the civilian-military dichotomy, that ‘in every society, at any given moment, there is a demarcation line considered as “normal” by the leading political forces’ (Bebler 1990:265). (Also see Nordlinger 1977:xii.)

[5] An early review of this literature is contained in Lowenthal (1974).

[6] More recent discussions of the ‘new professionalism’ of the military in Southeast Asia are contained in Soedjati and Yong (1988) and Selochan (1990).

[7] Some recent tendencies are discussed in Sarkesian (1981); Stepan (1988); Goodman, Mendelson and Rial (1990:Part III); Zagorski (1992); Burk (1993), and Ashkenazy (1994).

[8] Thus, although Finer observes that most military regimes have very short lives, he also notes that: ‘Few civilian successor regimes have lasted more than ten years’ (1985:29).

[9] Cf. Luckham (1991:10): ‘Rather than analysing coups as such, we might do better to consider them as part of a much wider process of transformation: firstly as a subcategory of a broader class of regime changes or political transitions; and secondly as one among several different channels through which military power can influence politics’.