22 Non Democratic Regimes authoritarian' is so widely applicable that it is difficult to develop a theory which can cover so many -diverse cases without becoming either banal or incoherent . In a general sense the term `authoritarian' could be said to describe a situation where (a) freedom is restricted in favour of obedience to authority, and (b) this authority is itself exercised with few restrictions (Schapiro, 1972: 39) . However, the notion of `authoritarian government' is often used as virtually a synonym for 'non- democratic government' . It is more widely applicable than the notion of dictatorship, which at least is not applied to monarchies and traditional forms of Xovernment. In fact the possible inclusion of tribal chiefs, priest-kings, and medieval monarchs as examples of authoritarian government has led some theorists of authoritarianism to redefine the term to specifically exclude such premodern forms of non-democratic government or regime . Some of these definitions of authoritarianism have also excluded totalitarianism, on the grounds that it is too extreme or distinctive a form of non-democratic government to be included in the same category with the more normal forms . Yet even when using these narrower definitions, theorists have been plagued by the problem of how to cover what is still a very diverse range of non-democratic regimes and forms of government . Even the classic works have either not been sufficiently coherent and systematic, or not been sufficiently broad and applicable . Linz on Authoritarianism Linz's pioneering 1964 analysis of authoritarianism, `An Authoritarian Regime : Spain', excluded totalitarianism as well as traditional monarchies and other traditional systems from his conception of authoritarianism (Linz, 1970 [1964] : 269-70) . But he rejected any notion that authoritarian regimes form only a residual category, such as the class of (modern) regimes that are neither democratic nor totalitarian . Instead, Linz stressed the distinctive nature of the authoritarian type of regime and presented a broad and multifaceted coverage of authoritarianism that was comparable to the theorieš `of totalitarianism (see Table 1 .3) . However, the signs of strain that are evident in this descriptive theory (and in his definition of authoritarianism) are evidence of how difficult it is to incorporate so many Theories of Non Democratic Government 23 varieties of non-democratic government into a single theory of authoritarianism . The prominence that Linz . gave to _military dictatorships in his description of authoritarianism highlighted the distinction between authoritarian and totalitarian regimes . Both the classic works on totalitarianism had noted that the military played a relatively minor role in a totalitarian dictatorship (Arendt, 1962 : 420; Friedrich and Brzezinski, 1961 : 273) ; in contrast, Linz pointed - out that the military enjoys a `privileged position' in most authoritarian regimes and that its position is likely to be further enhanced if the regime had been established by a military coup (1970: 267) . By also including party dictatorships within his conception of authoritarianism Linz provided a solution to the problem of how to classify less-than-totalitarian and post-totalitarian party regimes . The number of Third World party dictatorships (particularly the many one-party states created in Africa in the 1960s) which it had seemed inappropriate to classify as totalitarian could be classed as authoritarian, as could the many communist regimes that had passed through and beyond their (most) totalitarian phase . For Linz suggested that totalitarian regimes might appear more like some authoritarian regimes `if their = ideological impetus is weakened, apathy and privatization replace mobilisation, and bureaucracies and managers gain increasing independence from the party' (ibid.: 281) . But it is difficult to provide a concise and coherent definition of authoritarianism that covers both the military and party types of dictatorship and yet still distinguishes authoritarianíšťn from totalitarianism . Linz's definition pointed to four distinctive elements or features that define an authoritarian regime. Although they can be listed in similar form to totalitarian theorists' six-point syndrome or eight contours/ pillars (as in Table 1 .2), Linz's four defining elements or features of authoritarianism require some accompanying explanation : I . Presence of `limited, not responsible, political pluralism' (1970: 255-6). Linz viewed this limited political pluralism as the most distinctive feature of authoritarianism . The limits may be (a) severe or moderate, (b) legal or de facto, and (c) applied only to parties and political groups or to interest groups as well. But the crucial point is that there are groups which Theorist Examples Linz (1964) Military regimes and non-totalitarian party dictatorships O'Donnell (1973) Argentinian, on Bureaucratic Brazilian, Greek Authoritarianism and Spanish (as distinct from military regimes of traditional and 1960s and early populist types of 1970s authoritarianism) Origins 1 . Lack of 'popular consensus' under a democratic government, or actual organised political strife or even aborted revolution, or 2. Lack of political mobilisation of the masses by previous regime - e.g. by traditional monarchy, oligarchic democracy or colonial rulers Coup coalition of military and civilian technocrats confident they can solve economic and other social problems (see goals) TABLE 1 .3 Theories of authoritarianism Goals 1 . Political demobilisation or depoliticisation, or 2. Socially progressive or conservative policies that require some mass mobilisation 1. Solve economic problem: shift from import-substituting/ consumer-goods industrialisation to more 'intensive' industrialisation involving production of capital goods -~ socially 'painful' economic policies requiring political exclusion and deactivation of popular sector 2. solve political-social problem of polarisation between popular sector and propertied sector --~ political exclusion/ deactivation 3. solve political problem of 'mass praetorianism' -a political exclusion/ deactivation Structure 1 . Regimes without a party, or 2. Regimes with a party in both cases: (i) Individual leader or occasionally small group of rulers (collective leadership) (ii) Military usually has 'privileged position' 'Bureaucratic' in sense of key role played by public bureaucracies (military, etc .) and private bureaucracies but 'inconsequential' whether military govern Evolution Long-term: decline of any leader's charisma ; civilianisation of any military rule; decline in level of any political mobilisation present in early stages of regime but only rarely evolution into democratic regime Long-term tendency towards political isolation of coup/ ruling coalition and possible problems in attaining goals -r split in coalition over whether to withdraw from power - but any withdrawal leaves resulting democracy plagued by pre-coup problems 26 Non Democratic Regimes 2. are independent of the regime and have some political influence. At one extreme is General Franco's absolutist regime in Spain allowing independence and influence to the Catholic Church; at the other extreme is the officially liberal-democratic party dictatorship in Mexico actually encouraging some degree of political participation by a limited number of independently existing parties and groups . Absence of `elaborate and guiding ideology' and instead_ distinctive mentalities' (ibid.: 255-8) . Mentalities are apparently more emotional than rational and are not as future-oriented as the utopianism of ideologies . But Linz acknowledged that ideologies, were by no means unknown among authoritarian regimes (though more commonly found among party than military dictatorships), and that in fact an ideology may be loudly proclaimed by an authoritarian regime . 3. Absence of intensive or extensive `political mobilization' throughout most of a regime's history (ibid.: 255, 259) . Political mobilisation is the exception rather than the rule in the case of authoritarian dictatorships . The exception occurs - in the early stages of some authoritarian regimes during which there may be considerable and even very intensive (controlled) popular participation . 4. A `leader (or occasionally a small group) exercises power within formally ill-defined limits but actually quite predictable ones' (ibid. : 255) . Even when the regime's leader or leaders may seem to be absolutist, in practice this power is exercised within predictable limits rather than in a wholly discretionary or_arbitrary fashion . Linz refers to the military junta as an example of occasions where power is exercised by a small group of leaders, and presumably another example would be the party political committee, such as the Communist Party Politburo . That all four of these features are either quite complex or have significant exceptions is an indication of the problems involved in generalising about authoritarian regimes . However, Linz attempted to make his conception of authoritarianism more systematic by classifying the exceptions to the absence-ofmobilisation feature as a separate subtype . He described these mobilising exceptions as `populistic' regimes, whose level of mobilisation falls short of `the pervasiveness and intensity Theories of Non Democratic Government 27 of the_ totalitarian -model'- but is still quite exceptional when compared to the lack of political mobilisation usually found in authoritarian regimes (ibid. : 260) . These populistic regimes therefore form what is probably best described as the highermobilisation subtype ; the more usual examples of authoritarianism form their own subtype, comprising the lowmobilisation regimes. It appears at first glance that this distinction between subtypes also involves an important distinction in political structure the absence or presence of an official party . For, as the lowmobilisation subtype is described by Linz as includ* ~without parties, the higher-mobilisation subtype presumably comprises authoritarian regimes -that do have an official party (í'Iřriá.) (The without-pařťy/with-party distinction does not distinguish between military and party dictatorships ; there have beén many cases where a military regime has had an official party.) However, some with-party regimes clearly belong in the low-mobilisation subtype . And Linz suggested that some ofthe more populistic', mobilising one-party states in Africa might eventually experience a decline in the degree of political mobilisation - with their parties being transformed into patronage ráther than mobilising organisations (ibid.) . In fact he viewed such depoliticisation as characteristic of any `stabilized' authoritarian regime, with or without a party (ibid. : 259-60) . His wide-ranging account of regime evolution was more impressive than his typology of authoritarian regimes (ibid. : 269, 271-2, 280-1) . As well as the decline in levels of mobilisation, he identified several other trends in the long-term evolution of an authoritarian regime, including the decline of any charisma initially possessed by a leader; the institutionalising-- of the exercise of power through the development of general rules ; and the civilianisation of the military dictatorships . Yet, while an authoritarian regime might therefore undergo some considéráble changes over the long term, Linz noted that only rarely had such a regime evolved into a stable democracy . -L ňž also provided a wide-ranging, if somewhat complex, account of the origins of authoritarian regimes (ibid.: 260-1, 267) . The lack of popular consensus which he considered to be a precondition seems to occur in two very different situations: either 28 Non Democratic Regimes 1 . there has been a period of abortive revolutions, organised political strife, or simply lack of consensus (under a democratic government) ; or 2 . the preceding regime has not politically mobilised the masses, as when the preceding regime has been a colonial administration, a traditional monarchy or an oligarchic democracy . These situations in turn create two very different forms of opportunity for authoritarianism : 1 . an authoritarian depoliticisation of society that is `one way to reduce the tension in the society and achieve a minimum of re-integration' (ibid.: 261) ; 2. the masses are initially easy to manipulate by a (populistic) authoritarian regime because they have not previously been won over by any organised movement. Similarly, Linz identified two alternative and very different types of goals that have been pursued by authoritarian regimes (ibid.: 261-4) . Many of therri havesought an actual demobilisation or depoliticisatii -nf- šočiety and, consciously or unconsciously, have encouraged a process that Linz termed `privatization', in the sense of citizens shifting their attention from public affairs to private matters . In contrast, other authoritarian regimes have sought instead to implement socially progressive or conservatiye - policies that require a degree of mass mobilisation if they are t__ obe effirifndy, or effectively implemented. As with the relationship between regime structure and subtype, there seems to be no consistent connection between origins and_goals - a particular type of goal is not associated with a particular type of origin . It is true that in many cases an authoritarian regime that has originated in a period of political. or social strife has also had the goal of depoliticising_,or - demobilising its society. However, Linz's case study of authoritarianism - Franco's military-party regime in Spain - originated from the intense social and military strife of the Spanish Civil War, but initially sought to attain fascist-style social goals that required high levels of mobilisation. In the light of these anomalies and complexities,_it_._iss not surprising that Linz did not attempt to present his theory more systematically - such as by presenting it in the form of two Theories of Non Democratic Government 29 alternative subtypes of authoritarianism that differ from each other in origins, goals and structure . But the lack of a systematic and coherent framework may also explain why his multifaceted theory of authoritarianism has been much less widely used than his more concise (and seemingly straightforward) four-feature definition of authoritarianism . ODonnell's Bureaucratic Authoritarianism O'Donnell's 1973 classic work, Modernization and Bureaucratic Authoritarianism: Studies in South American Politi d L' ' definition of authoritarianism as the basis for a more narrowly focused theory that was aimed at a particular variety ofauťhoritarian regime, which he labelled `bureaucratic authoritarianism' -(see Table 1 .3) . O'Donnell viewed authoritarianism (as defined by Linz) as being the genus to which belonged three species of authoritarianism - the traditional; the populist and the bureaucratic types (O'Donnell, 1979 [1973] : 91, 91 n. 76) . O'Donnell's typology also differed from Linz's in being linked to levels of modernisation rather than mobilisation (ibid. : 108-9) . The traditional type of authoritarianism is associated with a 1ów level of modernisation, aš in his example of Stroessner's regime in Paraguay (see Chapter 6), but he specifically excludes any traditional monarchy or other traditional forms of government. The populist type of authoritarianism such as Peron s regime in Argentina (see Chapter 9), is associated with medium levels of modernisation . It seems somewhat similar to Linz's highermobilisation subtyp f the leaders of populist-authoritarian regimes attempt to politically activate and `incorporate' - under tight control - segments of the `popular sector', namely the working class and sections of the lower-middle class . O'Donnell was primarily concerned with the high-modernisation e of authoritarianism - the `bureaucratic' _type (ibid. : 90) .11 As he implies, this type is somewhat similar to Linz's depoliticising, low-mobilisation subtype because the bureaucratic type-seeks togolit callý -exclude áňď deactivate the popular sector . The description `bureaucratic' is not exactly self-explanatory but had been borrowed from Janos's recent analysis of non-democratic government in Eastern Europe in _the 1939s . Janos had argued thatin Yugoslavia, Hungary, Romania and Poland an `administrativemi nary complex' of civil service and military (together with 30 Non Democratic Regimes some middle-class camp followers) had formed a dominant political, class which he labelled 'bureaucratic' (Janos, 1970: 205) . Although the military component of O'Donnell's bureaucratic type seems particularly strong, this is largely because he used the military regimes established in Argentina and Brazil in the 1960s as his case studies or primary examples of bureaucratic authoritarianism . He specified that the presence or absence of military government was irrelevant t ypologically, and therefore bureaucratic authoritarianism can presumably also occur under a party dictatorship (O'Donnell, 1979: 108) . As for the description `bureaucratic', O'Donnell argued that the term suggested the typical features of the high-modernisation type of authoritarianism . These included not only the key role played by large public bureaucracies (military and civil service) and private bureaucracies (business corporations), but also the role played by technocrats, the organisational strength of many social sectors, and the government's attempts to control social sectors by `encapsulation' by encapsulating their political representation into government-dependent interest groups and/ or political parties (ibid. : 91, 51) . This encapsulation is in turn linked to the government's attempt to politically exclude and deactivate the popular sector . The government may (a) eliminate the sector's organisational bases so that `it can no longer make genuine political demands', and/or (b) eliminate the electoral arena by allowing `only government-sponsored parties to participate' or by simply suppressing all electoral activities (ibid.: 51-2, 52 n. 1) . Therefore O'Donnell had linked his high-modernisation subtype to a particular structure, the bureaucratic, and to_ a particular goal, that-of political exclusion and deactivation . But the depth as well as multifaceted nature of his account of bureaucratic authoritarianism is -most evident in his explanation of how political exclusion and deactivation is only an intermediate goal - a means to an end - that has to be attained in order_to achieve a wider and much more ambitious goal. For the bureaucratic- authoritarian regime's ultimate -goal is to solve major economic and political-social problems whičh had plaýed-álal role in the origins of the regime. The underlying problem was economic in nature and likely to be found in only highly modernised economies . Essentially the problem was that the country had exhausted the possibilities offered by the relatively easy, import-substituting/consumer-goods stage_ of industrial - growth, and was therefore suffering from a lack of economic growth, high inflation and foreign-exchange -problems (ibid. : 57-64) . To shift to a more `intensive' form of industrialisation (`the vertical integration of domestic industry for the production of a wide range of raw, intermediate, and capital goods') would require `quite painful' economic policies which could not be -implemented -unless there was a reduction in the popular sector's demands for consumer goods- -and - for' participation in political power (ibid. : 59-60, 63, 67-9) . The political exclusion and deactivation of the popular sector was the obvious authoritarian solution to this preliminary, political aspect of thececonomic problem . It was also the obvious authoritarian solution, even if only a temporary or stop-gap solution, to the two political-social problems facing the country. The popular sector's demands for consumer goods and political power had led to a social, classconflict `polarization' between the propertied sectors and the popular sector - a polarisation which had been intensified by the recent example of social revolution in Cuba (ibid. : 69) . The second political-social problem was a situation described by some political scientists as `mass praetorianism' (see also Chapter 2 and Exhibit 2.1), in which a society's political institutions are unable to cope with political participation by the ňibaň lower 'class (ibid. : 73) . The bureaucratic-authoritarian regime is established by a `coup coalition' of officers and civilians that intends to make and implement policies which will effectively deal with these problems (TRd_ 74, 81-5) . The coup coalition tends to be dominated by technocratic officers and civilians who share not only a commónJargon and approach, but also a common self-confidence in their capabilities, believing that `their combined expertise can ensure effective problem-solving throughout a broad range of social problems' (ibid. : 83) . O'Donnell also examined in some detail the evolution of these technocrats' bureaucratic-authoritarian regimes (ibid. 85, 105). Apparently the coup coalition can count on at least - the acquiescence of many sectors of society when it stages a coup and when it carries out its initial policy of political exclusion and deactivation . But this now ruling coalition (a) will become more isolated as it moves on to implement policies with high Theories of Non Democratic Government 31 32 Non Democratic Regimes social and economic costs, and (b) rnay only partly successful in implementing its exclusionary and economic policies . Such difficulties and failures will lead to a split within the ruling coalition over the issue of whether tress on or whether to withdraw from power and return the_country_ -to democracy. However, even if there is a withdrawal from power, the resulting democracy will be plagued by the pre-coup problems which bureaucratic-authoritarian government failed to solve . With his coverage of evolution, origins, goals and structure, O'Donnell had presented a multifaceted account of bureaucratic authoritarianism that was more coherent and systematic than Linz's general theory of authoritarianism . O'Donnell's theory was much more closely linked to the regime's characteristic feature - a high level of modernisation - and it also provided a much deeper and coherent account of the regime's origins. Therefore, it is not surprising that `bureaucratic authoritarianism' soon became one of the most widely known concepts in political science and one of the most widely used by analysts of South American non-democratic government . On the other hand, any theory focused on only one type of authoritarianism could be expectied to be more detailed and coherent than an attempt to cover authoritarianism in eneral. Moreover, by providing such a specific, detailed coverage of the bureaucratic-authoritarian regime's origins and goals, O'Donnell markedly reduced the applicability of his theory of high-modernisation authoritarianism . In fact the number of authoritarian regimes which meetthe origins/goalss criteria_ for bureaucratic authoritarianism may be as few as Arendt's examples of totalitarianism . O'Donnell's two examples from outside South America, the Spanish and Greek military regimes, do not seem to have been responding to this particular set of economic and political problems . The 1930s regimes in Eastern Europe from which the notion of `bureaucratic' regimes was first derived had not yet reached a high level of modernisation (Janos, 1970) . And none of the new high-modernisation dictatorships that emerged in South America in the 1970s seem to have followed the same path as O'Donnell's two examples. Although the notion of bureaucratic authoritarianism was the point of departure for a survey of-the new authoritarianism_in_ Latin America (Collier, 1979), not all of the contributors to Theories of Non Democratic Government 33 the book wholly accepted the bureaucratic-authoritarian `model', particularly its economic aspect . Doubts were raised about the link between bureaucratic-authoritarian policies and the high modernisation stage of economic development, and it was stressed that there was _no- objective' economic need for highmodernisation countries to adopt these political and economic policies . One contributor pointed out that such countries as Venezuela and Colombia had internationalised their production without the need for bureaucratic authoritarianism, and he contended that the economic successes of even O'Donnell's Brazilian example had not required an authoritarian government (Sera, 1979) . Another contributor (Kaufman, 1979) pointed to the Mexican case as evidence that some countries had achieved coruparable economic successes to O'Donnell's Brazilian and Argentinian examples without the need for bureaucratic authoritarianism. Kaufman also pointed to significant differences in economic policies among high-modernisation authoritarian regimes, and suggested that the common feature of such regimes' economic programmes was actually export diversification, not vertically integrated intensification or `deepening' of industrialisation . During the 1980s and into the 1990s, , the bureaucraticauthoritarian `model' faced a growing body of work criticising or discounting it. For example, it was shown to be seriously inapplicable to cases of personalise rule, such as General Pinochet's high-modernisation authoritarian regime in Chile (Remmer,_ 1989) . And its analysis of high-modernisation economic problems and goals continued to be discounted . For example, an analysis of the economic policies of the highmodernisation authoritarian regimes established in Argentina, Chile and Uruguay in the 1970s argued that they had implemented a `neoconservadve' programme of economic restructuring which was monetarist, efficiency-oriented, and free-market sharing `few features, if any, with the bureaucratic-authoritarian economic model' (Schamis, 1991 : 209-10) . Yet, the notion of bureaucratic authoritarianism was not deemed by conventional wisdom to be an outmoded concept, even jj°b gh its two model examples were defunct and no equivalents ever emerged in South America or elsewhere, and the term `bureaucratic authoritarianism' remained a convenient label to attach to high-modernisation dictatorships . For example, although Im (1987) acknowledged that the Park regime of 1970s South 34 Non Democratic Regimes Korea had a different economic, social and political background from the O'Donnell examples, this did not prevent him from labelling the regime as a case of bureaucratic authoritarianism . Therefore the notion of bureaucratic authoritarianism made an important contribution to the continuing prominence of the concept of authoritarianism, as reflected also in such works as Perlmutter's analysis of modern authoritarianism . Perlmutter on Authoritarianism Perlmutter's (1981) work, Modern Authoritananum: A Comparative Institutional Analysis differed markedly from Linž's - apprdach. Perlmutter did not exclude premodern forms of non-democratic government from his definition of `authoritarianism', preferring instead to draw a distinction between modern an o er forms by pointing out that the older regimes had seen rule by - the few in the name of the few, but that the modern regimes have instead seen rule by the few in the name of the many (ibid .: 2) . Perlmutter also differed from Linz (and reflected the increasing disillusionment with the notion of totalitarianism) by including the classic cases of totalitarianism, Nazi Germany and the Soviet Union, among his examples of modern authoritarianism . More importantly, Perlmutter emphasised the institutionalstructural aspect of authoritarian regimes, arguing that this was the most useful way of explaining their political behaviour and dynamics (ibid.: 62-3) . His structural analysis is not based, though, on the distinction between party and military regimes, but rather on the ways in which regimes employ particular structures or institutions as instruments for intervention, penetration and supervision These instruments are : 1 . `the single authoritarian party' ; 2. `the bureaucratic-military complex', which means basically the civil service and the military ; and 3. `the parallel and auxiliary structures of domination, mobilization, and control', such as political police, paramilitary forces, and militant youth movements (ibid. : 9, 11, 13) . Each of his many types or models of authoritarianism (such as the Bolshevik Communist, Nazi, Fascist, Corporatist and Praetorian models) apparently shows a distinctive or characTheories of Non Democratic Government 35 teristic preference for one of these three institutional-structural instruments. For example, one type or model may dispense `with the use of the single party and employ the military as the instrument of domination', while another may make `extensive use of the political police, an auxiliary instrument, at the expense of the single party and the military' (ibid.: 9) . Perlmutter had therefore rectifiedCa seribús weakness in Linz's and O'Donnell's_approaches,to authoritarianism - the lack of attention paid to structural features . It is true that he did not provide an alternative theory of authoritarianism : his analysis lacked sufficient coverage of origins and goals to be properly compared with Linz's and O'Donnell's theories, however flawed and problematic they may have been . But his less ambitious approach seems the only way in which the study of authontanamsm can adequately and coherently cover the various types o£ regime--tháthávégóve-(etheir- sóčietiěš in an áúthontarlan fashion. ť 2 Types of Non-Democratic Regime In this chapter attention will shift to less ambitious approaches to analysing non-democratic government . They focus on a particular type of modern non-democratic regime - military, party or personalist - rather than dealing with the wider topic of the form of government . As was pointed out in the previous chapter, theorists of non-democratic forms of government are primarily concerned with the question of how dictatorships rule, while analysts of particular types of non-democratic regime instead focus on the narrower question of who or what rules the military, a party or a personal leader . And these approaches are also often less ambitious in the sense of presenting only a typology, an analytic framework for classifying types and subtypes of regime, rather than presenting a descriptive theory comparable to those of totalitarianism and authoritarianism . However, the two classic works on the party type of nondemocratic regime, Tucker's and Huntington's, were exceptions that in addition to providing a typology, covered a similar range of topics to theories of totalitarianism and authoritarianism . Moreover, both Tucker and Huntington envisaged their respectively `single-party' and 'one-party' types of non-democratic regime as including not only cases of party rule, but also cases of military rule accompanied by an official party (which are actually cases of military rather than party dictatorship) . In contrast, analysts of the military type of non-democratic regime have more strictly applied the traditional criterion of `who rules?', and 36 Types of Non Democratic Regime 37 have included only cases of military rule - of the military's or a military man's dictatorship . But military rule has taken diverse forms that are sometimes far from straightforward . As an examination of Finer's classic structural typology of military regimes will show, military rule can take indirect and civilianised forms that are difficult to identify and/or to categorise . The many other typologies of military dictatorship have tended to be at least as concerned with the roles or goals of military regimes as with their varying structures, as will be seen in the case of Perlmutter's, Nordlinger's and Huntington's classic typologies of what they term `praetorianism' . Finally, any survey of types of modern non-democratic regime has to include the personalist type, with its personal rule by an individual leader rather than by his party or military . Personal rule has occurred in many modern dictatorships and is not confined to any particular variety, whether party or military, totalitarian or authoritarian, rightist or leftist . The presence of personal rule is usually viewed as being only a secondary or supplementary feature of a regime, not as a basis for classifying it as a personalist type of non-democratic regime . However, personalist types/ typologies have been developed by Linz and by Jackson and Rosberg . Moreover, the two personalist elements of Weber's venerable typology of legitimate rule, charisma and patrimonialism, have been used to categorise as well as analyse some non-democratic regimes . The Party Type of Non-Democratic Regime Tucker's Movement-Regimes' In his 1961 paper `Towards a Comparative Politics of MovementRegimes', Tucker provided a `user-friendly' typology of singleparty regimes that was based on the familiar, commonly used categories of communist, fascist and nationalist (see Figure 2.1) . He argued that the concept of totalitarianism needed to be supplemented by a wider category that would take into account the resemblances that communism and fascism shared with the `large and still growing number of revolutionary nationalist regimes under single party auspices' (1961 : 283) . He identified Kemalist Turkey, Nationalist China, Bourguiba's Tunisia, Nasser's 38 Non-Democratic Regimes FIGURE 2.1 Tucker's and Huntington's typologies of single-party or one-party regimes Tuckers genus/type Huntingtons type Revolutionary mass-movement regime One-party system under single-party auspices Species/subtypes Subtypes communist fascist nationalist revolutionary exclusionary established (evolves from revolutionary) Egypt, and Nkrumah's Ghana as a few of the many examples of single-party nationalist regimes . (As can be seen from his inclusion of the Turkish and Egyptian military regimes, Tucker was not concerned with whether the single party actually ruled or was only a subordinate part of the regime .) The nationalist 'single-party systems' were authoritarian rather than totalitarian, but they displayed sufficient similarities with communist and fascist regimes for Tucker to argue that the nationalist, communist and fascist regimes should be viewed as three species of the same political genus - which he labelled the `revolutionary mass-movement regime under single-party auspices' or, more concisely and confusingly, the 'movementregime' (ibid.) . Moreover, Tucker provided a multifaceted exposition of this typology, describing the movement-regime's structure, goals and evolution as well as subtypes . The single-party aspect of the movement-regime is a key component of the regime's structure (ibid. : 284-5, 288) . Not only does the regime operate under single-party auspices, but the party also controls the mass movement that mobilises popular participation in support of the regime and revolution - and therefore provides the regime with a claim to be democratic. He depicted the party as governing the country and dominating the regime, but later acknowledged that the party had a much weaker position in the two fascist regimes . In fact he labelled Types of Non Democratic Regime 39 them `fuehrerist' (leaderist) regimes because they were dominated by an autocratic/absolutist leader rather than by a party . The revolutionary goals of a movement-regime are to be found in its ideology (ibid. : 283-4) . The ideology contains the philosophy, programme and political orientation of the movement and its revolution, and after coming to power the movement attempts to maintain its regime's revolutionary momentum . Even a nationalist revolutionary movement seeks not only national independence as a sovereign state, but also the modernisation of this new nation-state. Such a goal `typically involves many elements of an internal social revolution' as old social relations and activities are `assailed in an internal revolution of national renewal' (ibid.: 286) . The presence of revolutionary goals does not mean that these regimes necessarily had revolutionary origins. For example, Nkrumah's movement came to power in Ghana through an electoral rather than revolutionary form of decolonisation, during which he won several elections and held the equivalent of a prime-ministerial post under British rule . His was only one of many African nationalist single-party regimes that originated in the 1950s-60s through electoral rather than revolutionary means. However, Tucker did not explore the issue of how revolutionary regimes could originate in a non-revolutionary fashion . In fact he was dubious about any attempts to account for why and how such regimes originated, contending that the movement-regime was a `political phenomenon to which no nation and no part of the world is immune' (ibid.) . However, Tucker did make some innovative points about the evolution of movement-regimes . He suggested that a movementregime could undergo a transformation, a `metamorphosis', from one species/type into another (ibid. : 289) . His example was the rise of Stalinism in the Soviet Union in 1928-38, which involved the metamorphosis of an originally communist movement-regime into a fuehrerist (fascist) type of movement-regime . More importantly (and convincingly), he argued that movementregimes which have lost their revolutionary momentum should be classed as `extinct' movement-regimes (ibid.: 286-7) . Although they may continue in power long after they have lost their revolutionary purpose, they will no longer be revolutionary regimes . It is not clear whether the revolutionarily `extinct' movementregime constitutes a new species/type, or even a whole new 40 Non-Democratic Regimes genus, of movement-regime, but some years after Tucker's work Huntington presented a more comprehensive typology/theory of party regimes that emphasised regime evolution and clearly specified that the fully evolved regime belonged to a new subtype . Huntington's Typology of One-Party Systems Huntington's typology appeared in his contribution to an edited work (Huntington and Moore, 1970) Authoritarian Politics in Modern Society : The Dynamics of Established One-Party Systems (see Figure 2.1) . He defined a one-party system as a political system in which there was only one effective party - any other parties that might exist having `little effect on the course of events' (1970: 5) . Like Tucker, he did not exclude cases where the party was attached to a military regime, as in Franco's Spain and Ataturk's Turkey. But his three subtypes described below were much less familiar than Tucker's communist, fascist and nationalist species/types of movement-regime . • The revolutionary one-party system . Of the Huntington typology's three subtypes, this is the most familiar, and apparently encompasses totalitarianism and Tucker's movement-regimes . It has the suitably revolutionary and extreme goal of seeking to liquidate or rapidly assimilate the politically subordinate section of its divided society (ibid.: 15) . Huntington describes this social division as a `bifurcation' in the sense of society being split into two sections on the basis of differences in class, race, religion or some other such category . The revolutionary type of one-party system includes the communist regimes, for they are seeking to eradicate class bifurcation by the elimination of the capitalist classes and the conversion of their former members into workers, peasants or intelligentsia. Huntington also categorised Nazi Germany as a revolutionary system, as the genocidal Nazi regime was seeking to eliminate the Jewish ethnic group rather than just exclude it from politics (ibid.: 17) . The African and other Third World one-party states apparently fall into the revolutionary category, but Huntington noted that the African systems were seeking to emphasise national unity and minimise sources of social conflict, such as the differences between modernised elite and traditional Types of Non Democratic Regime 41 masses (ibid.: 14) . Although he argued that this approach was a reason why the African examples were `weak' one-party systems, their emphasis on national unity did not affect their revolutionary credentials - as apparently the modernised elite was seeking to eradicate bifurcation by rapidly assimilating the traditional masses. • The exclusionary oneparty system. This second subtype is less familiar but, as was seen in Chapter 1, theories of authoritarianism have been very concerned with the issue of political exclusion or demobilisation. In Huntington's typology of oneparty systems, the exclusionary type was described as seeking to suppress or restrict the political activity of the politically subordinate section of its divided (bifurcated) society (ibid. : 15) . • Although this type is much rarer than the revolutionary, Huntington mentioned a number of cases of exclusionary one-party system : Liberia, South Africa, the South of the United States, Northern Ireland, Kemalist Turkey, Kuomintang China, and Taiwan . The exclusionary systems are more moderate than the revolutionary in other respects than just preferring suppression /restriction rather than liquidation /assimilation, and Huntington contended that it was inappropriate to apply the notions of totalitarianism or movement-regime to them . He also suggested that an exclusionary system will in the long term adopt one (or more?) of the following strategies : retreat into isolation from the world community, rein in economic and social change, become more repressive, or evolve into a form of competitive party system (ibid.: 18-23) . • The established one-party system . Huntington's third subtype differed fundamentally from the other two in being only an evolved form of revolutionary one-party system and in lacking any obvious social or political goal . The evolved, established one-party system is a complex regime that has several characteristic features (ibid.: 23, 40-1) . It is not faced with the issue of a divided, bifurcated society (which has been dealt with by the preceding, revolutionary one-party system), and it has a more administrative than revolutionary character . There has been a decline in party-mobilised popular participation and in the importance of ideology in shaping goals and policy decisions . Political leadership tends to have lost its personalist, charismatic and autocratic quality and to have become oligarchical and bureaucratic . The party 42 Non-Democratic Regimes elite is no longer an initiator of policy but rather a mediator of policy initiatives coming from technocratic and managerial elites. The party as an organisation has become the regulator of a now pluralist political and social structure containing important interest groups . Huntington used the uniquely liberal and decentralised version of communism developed in Yugoslavia as his example of this type, but other communist regimes had also journeyed far along the evolutionary path to an established type of one-party system . He provided an extensive analysis of how a revolutionary one-party system evolved into an established one-party system, and gave a detailed description of the three phases in this evolutionary process - transformation, consolidation and adaptation (ibid.: 24-38) . Huntington's extensive coverage of regime evolution was the main part of a multifaceted account of one-party systems which was comparable in breadth and depth to the descriptive theories of totalitarianism and authoritarianism . He had provided a brief coverage of regime goals when he had described the revolutionary and exclusionary types' different approaches to dealing with social divisions . And he also provided innovative analyses of both the origins and structure of the revolutionary and exclusionary types . The social origins of these one-party systems are closely linked to their goals, and are to be found in bifurcation between 'socialeconomic groups' or between `racial, religious or ethnic' social forces (ibid.: 11) . Moreover, it is crucial that these social bifurca tions be `pronounced' and `intense', as only `sharp' bifurcation will provide the basis for a viable one-party system (ibid.: 11-12) . Huntington pointed to tropical Africa as an example of how this sharp bifurcation is unlikely to arise at an early stage of modernisation, and he concluded that, therefore, any one-party system created in tropical Africa was likely `to be weak and fragile' (ibid.: 12) . As several African one-party systems had already fallen victim to military coups, he had some justification for questioning their viability and suggesting that they were likely to be fragile . Like Tucker, Huntington overemphasised the revolutionary origins of party regimes, contending that `the largest number of one-party systems are produced by social revolutions' (ibid.) . But most one-party systems had not been the product of social Types of Non Democratic Regime 43 revolutions ; most Third World one-party systems been the product of peaceful decolonisation or military coup, and the majority of communist one-party systems had been established with the assistance of the Soviet Union or through a war of liberation against foreign (fascist or colonial) occupation . However, Huntington argued that the strength of a one-party system was affected by the intensity and duration of the struggle to acquire power. In many African cases `the party came to power easily, without a major struggle' (that is, through a peaceful process of decolonising elections), and therefore the party `withered in power' (ibid. : 14) . As for Soviet-assisted cases of communist one-party systems, Huntington explained that these (seemingly easily established) party regimes had consolidated their power by intensifying class struggle in the initial, transformation phase of the regime . In his analysis of the structure of one-party systems Huntington distinguished strong from weak systems by considering two questions relating to the role of the party (ibid. : 6-7) . First, to what extent does the party monopolise (1) legitimation, (2) recruitment of leaders, and (3) policy-making and interestaggregation? Even in a strong one-party system the party never completely monopolises these three functions, and in a weak system the party may be relegated to only a minor role . Second, there was the question of whether other political actors, apart from the party, may play a significant and perhaps dominant role in the one-party system . Huntington identified five other, non-party types of political actor : personalistic, including a charismatic leader ; traditional, such as the church or monarchy (and presumably the tribe or ethnic group) ; bureaucratic, such as the civil service, police and military ; parliamentary; and functional socioeconomic, such as the working class, the peasantry, managers and intellectuals . In weak one-party systems one or more of these actors eclipse the party's role . In other cases there may be a balance of power between the party and other actors . And in strong one-party systems the party plays the dominant role. 1. 2. 3. 4. 5. 44 Non Democratic Regimes Brooker on Ideological One-Party States After Tucker's and Huntington's classic theories and typologies there was a familiar tendency for later analysts to be less ambitious in their approach . For example, Brooker's (1995) work on ideological one-party states was content to emphasise the distinction between the 'party-state' and 'military-party' subtypes, and to examine from a new perspective the role played by the official party in these two types of one-party state . He examined not only the extent to which the party performed the governing role of making policy and supervising its implementation, but also the extent to which the party performed political (electoral) and social (indoctrinating) roles . Moreover, he analysed in some depth the one-party or single-party nature of these regimes. The regime party's monopoly was seen as being (a) either a legal or a de facto monopoly in `literal' one-party states, or (b) an effective monopoly in `substantive' one-party states, in which other parties are allowed to exist but not to compete (successfully) with the official party . But there was still no move to define the party type of nondemocratic regime as including only specifically party dictatorships/regimes and excluding any military-party regimes . The Military Type of Non-Democratic Regime Finer on Military Regimes The many analysts who have presented typologies of the military type of non-democratic regime have left little doubt that they were analysing specifically military dictatorships, as their typologies have tended to focus on the form or structure of military rule and its role or goal (see Table 2 .1) . The tendency was evident even in such early works as Finer's 1962 The Man on Horseback : The Role of the Military in Politics, whose classic typology identified five different forms of military rule and regime : two direct, two indirect and one dual (1976 [1962] : 149-51, 245-6) . The more standard of the two forms TABLE 2.1 Typologies of military dictatorship Types of Non Democratic Regime 45 of direct rule involves openly military rule by a military junta or by a military government, with leading military officers installed as the country's president and/or government ministers . The `quasi-civilianised' form of direct military rule differs by clothing itself in (supposed) evidence of civilian support, and/ or in civilian garb and institutions (ibid.: 163) . Its civilian features may even include a supportive political party, but all the regime's civilian institutions are only `civilian trappings, emanating from and dependent on the military' (ibid.: 159) . In contrast, the civilian component of the `dual' type of military regime - a political party or some other organised civilian support - has been developed by a military dictator as reliable `civilian forces' that can act as `a counterpoise to the views and the influence of the army' (ibid. : 150, 158) . As the head of both the military and this civilian organisation he can strengthen his personal position by establishing a balance of power between the dual, military and civilian bases of the regime . Finer Structural forms of military rule Perlmutter Goals/roles and structure Nordlinger Goals/roles and structure Huntington Goals/roles 1 . Indirect- 1 . Arbitrator 1 . Moderator 1 . Oligarchical limited 2 . Ruler (limited goals, (shift to 2. Indirect- 3. Party-army indirect rule) radical) complete (evolves from 2. Guardian praetorian 3. Dual ruler) (limited goals, society with 4. Direct direct rule) radical/ 5. Direct : quasi- 3. Ruler reformer civilianised (ambitious military goals, 2. Middle-class extensive radical rule) praetorian society with arbiter/ stabiliser military 3. Mass praetorian society with guardian/ vetoer military 46 Non Democratic Regimes Finer's typology also emphasised how military rule can take the indirect form of controlling á civilian government from behind the scenes (ibid.: 151-7) . He identified two types of indirect military rule and regime : 1. the 'indirect-limited' type sees the military exerting control over the government only intermittently and to secure only limited objectives ; 2 . the 'indirect-complete' type sees the military continuously control all the activities of the civilian government . In 1981 Finer returned to the issue of categorising military regimes and added some new features to his classic typology (1988: 255-72) . Now he focused on the question of the extent to which the military `as such' takes on a governing or policymaking role in a military regime . Here he seems to have been drawing the distinction between (a) rule by the military as an organisation, and (b) rule by military men operating as free agents or personal rulers . Surveying the whole field of thenexisting military regimes, he looked in each case for the presence of a military junta and/or cabinet, as he assumed that this feature was an indication of rule by the military `as such' - at least to the extent that supreme executive power is wielded by military men `who in some sense or other command and/or represent the armed forces' (1988 : 255) . He noted that such regimes could be considered as belonging to the `direct' or the `direct : quasi-civilianised' categories of his long-established typology . However, Finer did not suggest how to categorise the many regimes that lacked a military junta/cabinet but were ruled by a military president . He viewed them as possibly cases of the military (as such) playing only a supportive rather than a policymaking role - supporting personal-presidential government by a military man. As a number of these regimes have an official party, they seem quite similar to Finer's long-established `dual' type. But he clearly created a new category of regime when he referred to the military playing a supportive and vital role in 'military-supportive' civilian regimes, which arise when a civilian government has to rely on active military support for its survival but is in no way a puppet of the military . In his 1962 work, Finer had also made a pioneering distinction relating to the ruling military's role rather than structure. 2. Perlmutter on Military Regimes A very similar distinction, between arbitrator and ruler types of ruling or `praetorian' army, was made by Perlmutter (1974) and later included in his wide-ranging (1977) work on professional, praetorian and revolutionary armies, The Military and Politics in Modern Times. The main characteristics of Perlmutter's `arbitrator' type are : '(1) acceptance of the existing social order ; (2) willingness to return to the barracks after disputes are settled; (3) no independent political organization and a lack of desire to maximize army rule' (1977 : 104-5) . His `ruler' type's main characteristics are : `(1) rejects the existing order and challenges itsolegitimacy; (2) lacks confidence in civilian rule and has no expectation of returning to the barracks ; (3) has a political organization and tends to maximize army rule' (ibid.: 107-8) . It has two subtypes that are again defined by the military's role or goal: the (radical or reforming) antitraditionalist subtype and the conservative antiradical subtype . Perlmutter also identified and described another type of military regime - the 'party-army' regime - that would have to be added to his arbitrator /ruler typology (ibid. : 145-7) . The obvious structural characteristic of this type is the presence of a political party, whether created or taken over by the military . (A ruler-type regime, too, may have a party, but may instead have a highly political kind of junta, such as a Revolutionary Command Council .) And in fact the party-army type differs from the other two - arbitrator and ruler - in being defined according to its structure rather than its goals . Evolving from a ruler-type regime, it sees a politically neutralised military return to the barracks and leave the regime in the hands of a military leader who has an official party at his disposal (in a situation reminiscent of Finer's dual-type regime) . Types of Non Democratic Regime 47 He pointed to two quite different ways in which the ruling military may conceive of its duty of custodianship of the national interest : 1. as arbitrating or vetoing of civilian political affairs that threaten the national interest, or as requiring `overt rulership of the nation and the establishment of a more or less complete political programme under their [the military's] authority' (1976 : 31) . 48 Non Democratic Regimes Nordlinger on Military Regimes Nordlinger's (1977) Soldiers in Politics : Military Coups and Governments appeared in the same year as Perlmutter's work and presented a typology of `praetorianism' that contained yet another example of the ruler(ship) category . In Nordlinger's typology the military's power reaches its greatest extent in this `ruler' type of regime . Here the military `not only control the government but dominate the regime, sometimes attempting to control large slices of political, economic and social life through the creation of mobilization structures' (1977 : 24) . This extensive control over state and society arises from the need to make far-reaching changes in order to attain the ruler type's ambitious political/ economic goals. As these ambitious goals also require the military to stay in power for an indefinite period, Nordlinger's version of the ruler type of military regime seems very like Perlmutter's . However, Nordlinger also included the important point that ruler-type regimes were relatively rare constituting `roughly 10 percent of all cases of military intervention' (ibid. : 26) . Nordlinger's typology of praetorianism - the ruler, guardian and moderator types - was unusually systematic . His types were defined by a combination of two variables: (a) the extent of a regime's political/ economic objectives or goals, and (b) the extent of the governmental power wielded by the military (ibid.: 22-6) . The ruler type is distinguished from the other two types both by its more ambitious goals and by the extensiveness of the power wielded by the military . The other two types have quite limited goals, but the moderator type is distinguished from the guardian type by its preference for indirect rather than direct rule . Although the goals of the guardian and moderator types were of an arbitrating/vetoing nature, they seem more reminiscent of Huntington's classic analysis of praetorian regimes than of Finer's work (see below and Table 2 .1) . The guardian type bears a marked resemblance to Huntington's conception of the vetoing, guardian role performed by the military in a `mass praetorian' situation, and of the arbiter/ stabilising role performed by the military in a `middle-class radical praetorian' situation . Hunting-ton on Military Regimes Huntington's (1968) classic work on Political Order in Changing Societies included what has become the most famous account of the roles performed by military regimes . However, the book's theory and typology of `praetorianism' actually referred to praetorian societies rather than regimes (see Exhibit 2.1) . The three different types of praetorian society are each associated with a particular level of political participation : the oligarchical type is associated with a low level of political participation, the middle-class radical type with a medium level, and the mass type with a high level of participation (ibid. : 80) . And in each type of praetorian society the military performs a distinctive and `typical' role . The oligarchical type of praetorian society is the oldest and least complex, predominating in nineteenth century Latin America and being common in the Middle East in the midtwentieth century (ibid. : 199-201) . It is associated with a low, traditional level of political participation that is limited largely to the dominant social forces - the big landowners, and the leading figures in the clergy and the military. The dominant political figure may well be a military man, but he is usually a highly personalist leader who is not the leader of the military as an institution - in fact the military lacks any autonomous existence . However, the military plays an autonomous and also vital role in the shift from the oligarchical to the middle-class radical type of praetorian society, which arises from a `breakthrough' or `reform' coup by (usually) middle-ranking officers (ibid.: 201-5, 222, 214, 209) . The officers who overthrow the oligarchical regime normally come to power with a programme of reforms aimed at achieving national integration and development, social and economic reform, and some extension of political participation . As it is the middle classes who benefit from the extension of participation, the middle-ranking officers leading the coup are depicted by Huntington as being the vanguard of the middle class, spearheading its `breakthrough' into the political arena . (Huntington used a wide interpretation of middle-class that included union-organised industrial workers .) By the midtwentieth century most praetorian societies in Asia, Africa and Latin America were of the middle-class radical type . Types of Non Democratic Regime 49 50 Non Democratic Regimes Exhibit 2 .1 The Praetorian Society The term 'praetorian' is derived from the historically famous Praetorian Guard units of the ancient Roman army, who exploited their position as guardians of the Emperor and capital city to put their favoured candidates on the Imperial throne . By analogy, the term 'praetorianism' has long been applied by political scientists to a situation of chronic military intervention in politics, or of the military exercising independent political power (Nordlinger, 1977: 2-3; Perlmutter, 1977: 90-3). But in 1968 Huntington gave the term a much wider meaning by using it to describe a type of society or polity in which the military are only one of the groups resorting to such direct action as staging coups ; where military intervention is only one particular manifestation of a general politicisation of social forces, groups and institutions (1968: 194-6). He argued that societies with a politicised military also have a politicised clergy and civil service and politicised universities, trade unions and business corporations . The political scene is plagued by a variety of forms of direct action as each social group uses its distinctive means of exerting direct political pressure : businesses bribe, workers strike, students riot and the military stage coups . It was presumably because military coups are 'more dramatic and effective' than other groups' means of direct action that Huntington felt justified in labelling ('for the sake of brevity') such societies as 'praetorian' (ibid.: 195-6) . Huntington also argued that modernising societies tend to be praetorian because their political institutions are too weak to handle the increasing levels of political participation that accompany modernisation (ibid.: 79-80). A praetorian society lacks 'effective' political institutions that are 'capable of mediating, refining, and moderating group political action' (ibid.: 196) . And, if political institutions are to be effective, they must keep pace with increases in the level of political participation, which in turn increases as a society is modernised - from a low level of participation in traditional society to medium levels in transitional societies and, finally, to high levels in modern societies . Similarly, the failure of political institutions produces a different type of praetorian society according to the level of participation : oligarchical (low), middleclass radical (medium), and mass (high) . Types of Non Democratic Regime 51 The military play a prominent but largely reactive, `arbitral or stabilizing' role within a well-established middle-class radical praetorian society (ibid.: 216-18, 222) . Military intervention is usually in response to escalating social conflict between other groups and `serves to halt the rapid mobilization of social forces into politics and into the streets' (ibid.: 216) . The military is the only social group or force that can take over the role of the increasingly ineffective government; only the military has some capacity to govern and to restore order - at least temporarily - by producing a rapid political demobilisation . After the shift from the radical to the mass type of praetorian society the military's role still tends to be reactive and politically demobilising . But now it plays a `guardian' role on behalf of the middle class, employing a `veto' coup to protect the middle class's dominant position against the now politically participating lower classes (ibid. : 222-3) . Mass praetorianism sees an extension of political participation to include the lower classes - such as urban unorganised labourers living in slums and shanty/squatter settlements - but the military seeks `to block the lower classes from scaling the heights of political power' (ibid. : 222) . However, Huntington's depiction of the three types of praetorian society (and the political roles of the military) has not proved to be widely applicable . His account is clearly based on the experience of Latin America, and he himself recognised that other regions of the world have had a very different political history (ibid.: 199-200) . For example, he acknowledged that the military-led breakthrough to radical praetorianism may involve the overthrow of a traditional monarchy (as in the Middle East) rather than the replacing of oligarchical praetorianism . He also acknowledged that there were major differences between the Latin American and African situations (ibid.) . The much more recent timing of decolonisation in Africa and the less stratified structure of society meant that African countries had not experienced oligarchical praetorianism . More importantly, in Africa it was civilian nationalists, not the military, who had led the middle-class's breakthrough to political participation (during the process of decolonisation) . When the middle-ranking military officers took power in Africa, they were removing middle-class nationalist governments that Huntington could accuse only of lacking legitimacy and authority, not of 52 NonDemocratic Regimes failing to cope with escalating social conflict . Yet, despite its various failings and flaws, Huntington's concept and depiction of praetorianism remains one of the most striking and famous contributions to the study of military rule . The Personalist Type of Non-Democratic Regime Weber's Typology The most influential typology of personalist rule has long been the historical-sociological analysis of types of legitimate rule that Weber had developed not long before his death in 1920. None of Weber's three types - the traditional, the charismatic and the legal-rational - was labelled `personal' or `personalist', but his typology was very concerned with the personalist aspect of rule . He pointed to how the impersonal nature of the legal-rational type of legitimate rule contrasted with the personalist nature of the other two types . In the case of the patrimonial subtype of traditional legitimacy, `obedience is owed to the person of the chief or monarch, and in the case of charismatic legitimacy the leader is `obeyed by virtue of personal trust in him' (Weber, 1964 [1922] : 328) . Moreover, so many later analysts and researchers have applied these parts of Weber's typology to personalist dictatorships that his notions of `charismatic' and `patrimonial' rule have become two of the best-known terms used in the study of non-democratic regimes . Weber viewed charisma as `a certain quality of an individual personality by virtue of which he is set apart from ordinary men and treated as endowed with supernatural, superhuman, or at least specifically exceptional powers and qualities' (ibid. : 358) . The charismatic type of rule is therefore highly personal (ibid.: 358-60, 364) . A charismatic leader, such as a religious prophet or a heroic warrior, is the recipient of a wholly personal devotion from those who recognise his charismatic quality and mission, and after his death his successors have to `routinise' this charismatic rule into the traditional or legal-rational type . (Even during his own lifetime the charismatic leader must continually provide proof, by means of miraculous successes, that he retains his charismatic quality ; if success should for long desert him, then so will his followers .) Although Weber did Types of NonDemocratic Regime 53 FIGURE 2 .2 Weber's typology of legitimate rule (with modern adaptations) Types Traditional Charismatic Legal-rational (personal) (personal) (impersonal) patrimonial subtype neopatrimonial personal rulership (modern (Roth's modern adaptation) adaptation) sultanism (Linz's modern adaptation) not envisage the rise of charismatic dictators, he confirmed that charisma can appear within a modern political context, mentioning demagogic intellectuals and charismatic party leaders . Ideological `prophets' and political heroes are therefore perhaps modern political equivalents of religious prophets and heroic warrior leaders . However, these are problems in applying the concept of charismatic legitimacy to modern dictatorships, despite the number of cases where a dictator has been claimed by his propagandists to possess exceptional personal qualities as political hero, ideological prophet and so forth . This `personality cult', as communist detractors of Stalin termed the glorification of a regime's personal leader, has arisen in many party dictatorships and even a few military regimes . But judging the extent to which the claim has been recognised by the mass of purported `followers' is virtually impossible . For recognition of a claim to charisma is a voluntary, internal acceptance which is impossible to judge accurately in a regime where public expressions of personal devotion to the leader are required on pain of dismissal, arrest, torture, imprisonment and even death . There are cases where the credibility of the claims and the apparent spontaneity of seemingly unfeigned displays of devotion can lead outsiders to confirm that charismatic legitimacy exists . But these judgements, 54 Non Democratic Regimes such as in the case of Egyptian leader Nasser's charisma, are still likely to be controversial and problematic (Dekmejian, 1971 and 1976 ; Bowie, 1976) . Weber viewed patrimonialism as only a subtype of the traditional type of legitimate rule . The two other subtypes, patriarchialism and gerontocracy, involve rule (by patriarch or elders) `on behalf of the group as a whole' (Weber, 1964 : 346) . In contrast, the patrimonial chief rules over `subjects' rather than fellowmembers of the group - he exercises `personal authority, which he appropriates in the same way as he would any ordinary object of possession' (ibid.: 347) . Tradition limits as well as legitimates the use of these personally appropriated public powers, but the patrimonial ruler is allowed a degree or sphere of arbitrariness that in some societies reaches an absolutist extreme which Weber terms `sultanism' (ibid.) . Moder n Applications of the Concept of Patrimonialism Tradition and patrimonialism Weber's classification of patrimonialism as a type of traditional rule has complicated attempts to apply the concept to modern regimes. For example, at the close of his study of West African one-party states, Zolberg (1966) argued that they approximated Weber's patrimonial type in some important respects, but questioned whether it was possible to speak of first-generation regimes being based on traditional legitimacy. He therefore loosened and expanded the concept of traditional legitimacy to mean a legitimacy based either on the notion `this is how things have always been', or on what he termed a 'past-orientation' - such as the official party's past glories (ibid.: 143-5) . Patrimonial Personal r uleInstead of revamping the notion of traditional legitimacy, Roth (1968) separated the notion of patrimonialism from that of traditional legitimacy and emphasised the personalist aspect of patrimonialism . This `detraditionalized, personalized patrimonialism' was (somewhat confusingly) labelled `personal rulership' and was said to be based on personal loyalties `linked to material incentives and rewards' (1968 : 195-6) . But Roth still referred to Weber's archaic term `sultanism' when describing the `highly centralized variant of personal governance' that allows the ruler `maximum discretion' (ibid. : 203) . The Types of Non Democratic Regime 55 influence of Weber was also evident when Roth contended that the concept of (patrimonial) personal rule was best applied to the new states of the Third World - some of which `may not be states at all but merely private governments of those powerful enough to rule' (ibid. : 198, 196) . A virtual typology of Roth's concept of (patrimonial) personal rule was later developed by Linz (1975) to categorise some non-democratic regimes that seemed quite distinct from totalitarian and authoritarian regimes (ibid. : 253, 259-64) . Adopting Roth's notion that personal rule is based on personal loyalty linked to material rewards, Linz identified four different systems of personal rule : modern sultanism, oligarchic democracy, caudillismo (rule by military chieftains), and caciquismo (rule by local political bosses) . Modern sultanism was the most obviously personalist of these four types of personal rule, and Linz's conception of modern sultanism could be used to describe a type of personalist dictatorship that did not readily fit into other categories . For example, the dictatorship established by `Papa Doc' Duvalier in Haiti in the late 1950s had been based upon neither the military nor a party but instead on a paramilitary/ political organisation, the Volunteers for National Security or `Tontons Macoutes' (bogey men), which provided him with a body of local political loyalists and terroristic political police (Ferguson, 1988) . Linz considered the modern sultanist system to be relatively rare and most likely to be found in small countries with largely agrarian economies and few urban centres, such as Duvalier's Haiti, Trujillo's Dominican Republic and other Central American countries (Linz, 1975 : 253, 259-62) . He followed Roth and Weber in viewing this sultanistic rule as the most centralised and most arbitrary or discretionary form of personal rule . Other important similarities with traditional patrimonialism are the private use of public power (as exemplified by corruption), and the personal nature of the ruler's staff - which tends to include his family, friends, cronies and even business associates. Linz acknowledged that sultanist personal rule was based on the extensive use of fear as well as on personal loyalty linked to material rewards. The army and the police `play a prominent role' in the sultanist regime, and `men directly involved in the use of violence to sustain the regime' are members of the ruler's personal staff (ibid.: 260) . But, unlike in totalitarian regimes, 56 Non Democratic Regimes the use of terror is not ideologically motivated or justified, and the regime's lack of any ideological commitment is also evident in the lack of a mass party, the absence of mass mobilisation, and the ruler's enriching of himself, his family and other members of the ruling group (ibid.: 217, 260, 189, 262) . Neopatrimonialism and Patrimonialism The term `neopatrimonial' (Eisenstadt, 1973) was to prove more popular than Roth's no= tion of `personal rule' in drawing the distinction between traditional and modern patrimonialism . The concept of neopatrimonialism seemed particularly applicable to African states, and was further developed for this purpose in the 1980s and 1990s (Medard, 1982; Van de Walle, 1994), but it was also applied to such non-African regimes as Pinochet's military dictatorship in Chile (Remmer, 1989) . Other researchers preferred to apply the concept of traditional patrimonialism to a modern regime rather than use some concept of neopatrimonialism. For example, Crouch (1979) viewed the patrimonial features of Suharto's Indonesia in terms of the persistence or reemergence of traditional features dating back to the precolonial Javanese kingdoms. The common element in these various uses of the older or newer concept was that characteristic features of patrimonialism (recast in suitably modern form) were viewed as important features of a contemporary regime that was based upon modern organisations and, officially or publicly, upon modern bureaucratic and legal norms. Other Conceptions of Personal Rule Not all conceptions of personal rule have been derived from Weber's typology. Jackson and Rosberg's (1982a) Personal Rule in Black Africa was influenced by Weber but was based on the older distinction between government (a) by individuals, and (b) by laws and institutions (ibid. : 9-10) . In their conception of personal rule, the absence of effective political institutions results in the predominance of a personal authority and power that is limited more by countervailing power than by institu tions; personal rule is a system of personal relations centred on the ruler's relations with his associates, clients, supporters and rivals (ibid.: 12-19) . Such personal rule is also `inherently Types of Non Democratic Regime 57 authoritarian' in its monopolising of power (denying political rights and opportunities to competitors), and it can arise in military as well as party regimes (ibid. : 23, 37-8) . Jackson and Rosberg's descriptive theory of African personal rule included a typology comprising the prophetic, the princely, the autocratic and the tyrannical types (ibid. : 77-81) . Prophetic personal rulers are visionaries who place great emphasis upon ideology and seek to shape society to fit the principles and goals of the ideology - usually a form of African socialism . In contrast, the princely type of personal ruler tends to base his rule on patron-client relations and manipulation rather than ideology and to rule in alliance with other oligarchs . The autocratic type of personal ruler, though, tends not to share power with, other leaders . His unlimited discretionary power is reminiscent of absolute monarchy and of Weber's notion of traditional sultanist patrimonial rule . The tyrannical type of personal ruler is similar to Linz's notion of modern sultanist personal rule, as the tyrannical personal ruler exercises a wholly arbitrary power that is based upon both instilling fear in his subjects and rewarding his collaborators, who are kept personally dependent upon him . This typology could also be applied to many non-African regimes. Jackson and Rosberg noted that personal authoritarianism had existed in modern non-African states, such as Mussolini's Italy and Franco's Spain, and that `marked features' of personal rule had been exhibited by Hitler's Germany and Tito's communist Yugoslavia (ibid. : 21) . Moreover, the typology could also be applied in a supplementary fashion to distinguish between different types of military or party dictatorship - and, used in this way, it highlights important differences between not only personal and non-personal but also different types of personal military or party dictatorship . Assessing the degr eeof personal rule is another important way in which the concept can be used in analyses of party or military dictatorships . The existence of varying degrees of personal rule was implied in Jackson and Rosberg's reference to `marked features' of personal rule being present in some non-African states. The existence of degrees of personal rule was more explicitly recognised in Brooker's description of the degeneration from `organisational' (party or military) to personal dictatorship (1995 : 9-10, 18) . He suggested that in cases 58 Non Democratic Regimes of highly personal rule, the initial principal-agent relationship between ruling organisation and regime leader may be actually reversed, with the organisation now only an agent of the personal ruler. Although this organisation-individual, principalagent approach to personal rule lacked the depth and breadth of Jackson and Rosberg's theory, it enabled him to include an assessment of (the degree of) personal rule in his analysis of one-party states . 3 The Emergence of Military Dictatorships The military dictatorship has been the most common form of dictatorship in modern times . Between the end of the Second World War and the onset of the 1970s-90s wave of democratisation, the military had `intervened in approximately two-thirds of the more than one hundred non-Western states', and in the later 1970s `controlled the government in about one-third of these countries' (Nordlinger, 1977 : 6) . It is not surprising, therefore, that many attempts have been made to explain or analyse why and how military dictatorships emerge . Yet the framework developed by Finer in the early 1960s still offers the most comprehensive approach to analysing the military intervention in politics which results in the emergence of military regimes. Finer argued that the military's intervention in politics, such as when a military coup overthrows the civilian government, is best explained by examining both subjective and objective factors - what he termed the disposition and the opportunity to intervene in politics (1976 [1962] : 74) . He presented the relationship between disposition and opportunity as a `calculus of intervention' . This calculus does not resemble the differential or integral calculus of mathematics, but rather is a form of political calculation (in the sense of a `calculating' politician) which may be very similar to the calculations of officers contemplating a military coup . The calculus of intervention is derived by combining the two variables, disposition and opportunity, Ko