8
Democratisation
The Demise of Dictatorships and Birth of Democracies
Dictatorships come to an end in various ways. They can be ended by such 'accidental' events as the death of a personalist dictator, or a foreign invasion that conquers the regime's country. More purposefully, a regime leader may end a party's rule by misappropriating power (see Chapter 6), or the party's expropriation may instead arise from a military coup or a spontaneous revolution by the people. But there have also been many cases; especially among military regimes, of relatively voluntary handing over of power to the people - a relinquishing of public offices and powers.
Relinquishing involves a voluntary transferral of power rather than a biological loss of power through death in office, or a political loss of power through being expropriated by seizure or misappropriation. However, various forms of force or pressure, such as armed insurrection or demonstrations of popular discontent, may contribute to the decision to relinquish power, if only by altering the regime's assessment of whether holding power is worth the effort or odium. Therefore relinquishing power involves degrees of 'voluntariness' and in extreme cases may mean little more than making the best of a virtually hopeless situation, with the regime's opponents likely to seize power in the near future. Such a situation comes very close to being a case of orderly expropriation rather than relinquishment -of being a surrender rather than a transferral. But in cases of orderly expropriation, in which the seizure of power ends in surrender negotiations and agreements, the regime has already
188
Democratisation 189
lost the use of its public powers - it has lost control of the streets and of society - and its public offices are all it has left to surrender.
Whether ended by death, expropriation or relinquishment, a dictatorship has often been replaced by simply a different type of non-democratic regime. The death of a personalist ruler has often resulted in a return to the preceding party or military type of dictatorship. And the ending of a party or personalist dictatorship by military coup has often resulted in the establishment of a military regime.
However, in recent decades democratisation has become the most common result of a dictatorship's demise-. Democracy, even if only short-lived, has always been a common ending of military dictatorships, but it became the common ending of party and personalist dictatorships, too, when most of the existing examples of these regimes were democratised in the 1980s, and early 1990s as part of what has been termed history's 'third wave' of democratisation.
The Third Wave of Democratisation
In the early 1990s Huntington used this term to describe the wave of democratisation that was still sweeping through the world and which he viewed as having begun in 1974 with the pro-democracy military coup that overthrew Portugal's longstanding dictatorship (Huntington, 1991: 3). He referred to it as the third wave to distinguish it from the two earlier periods of relatively frequent transitions to democracy: the first, 'long' wave of democratisation in 1828-1926, and the second, 'short' wave in 1943-62 (ibid.: 16). He pointed out that in 1990 the actual proportion of democratic states in the world (45 per cent) was still no higher than its previous peak in 1922 (ibid.: 26). But the democratisation of numerous states in Africa during the early 1990s would give the democratic states a large majority - compared to being very much in the minority (25 per cent) before the onset of the third wave of democratisation.
Huntington also made an impressive attempt at explaining why this wave of democratisation occurred. He suggested that five 'independent variables' played significant roles in bringing it about:
190   Non-Democratic Regimes
1. authoritarian regimes' increasing legitimacy problems, including problems with (economic) performance legitimacy;
2. the social effects of the 1960s rapid economic growth, such as the expansion of the middle classes;
3. changes in the political attitudes of the Catholic Church, which nationally and internationally became opposed to authoritarian regimes;
4. changes in external actors' policies, such as the Soviet Union abandoning its policy of intervening militarily to maintain communist rule in Eastern Europe; and
5. the 'snowball' or 'demonstration' effects (see Exhibit 8.1), enhanced by increasingly effective international mass media, as earlier transitions in the wave of democratisation stimulated people in other countries to attempt similar regime changes (Huntington, 1991: 44-5).
However, explanations or explanatory theories of democratisation have their limitations when used to analyse the ending of dictatorship - the expropriation or relinquishing of power - that has led to the birth of democracy. They seldom provide a systematic analytical framework that (a) focuses on a few crucial elements in the ending of a regime, such as the strength of opposition to the regime, and (b) defines these elements in sufficiently broad terms to incorporate a wide range of more specific, explanatory factors - such as the changes in the political attitudes of the Catholic Church and the 'snowball' or 'demonstration' effects of democratisation.
Another limitation of theories of democratisation is that they usually do not explain important regional or regime-type variations in the way that dictatorships are democratised and in their vulnerability to a wave of democratisation. Usually theories of democratisation have to be supplemented by explanations of democratisation in a particular region or in relation to a particular type of dictatorship. (The regional explanations may take the form of theories explaining why democratisation occurred in a particular region, such as in Latin America or Africa, or they may take the less explicit form of a general or global theory including explanations for regional variations.) The need for regional explanations can be seen in the regionally 'biased' manner in which the 1970s-90s wave of democratisation swept through the world. The lack of democratisation in the Middle
j'll J, Democratisation 191
ill1; il
Exhibit 8.1   The Domino, Snowball or Demonstration Effect
An important factor in transitions from dictatorship to democracy is popularly known as the 'domino' or 'snowball' effect. The 'domino' metaphor had originally been used in Cold War rhetoric to highlight the danger that communist revolution would spread rapidly from one Asian country to another. By the 1980s, though, it was 'democratic dominoes' that were increasingly evident as many military regimes in Latin America and Asia succumbed to democratisation. Starr's (1991) statistical analysis of 1974-87 democratisation - a period in which transitions were predominantly from military rule - concluded that such a 'domino' effect did exist, though domestic factors made a country 'ready' for democratisation or produced 'barriers' against this external democratising effect.
Moreover, his analysis indicated that the domino or snowball effect wais evident regionally as well as globally, as it concluded that a powerful regional effect was evident in Latin American democratisation-in 1984-86. The standard technical term for such a global and/or regional phenomenon is 'diffusion effect', but it is also referred to as an 'emulation', 'modelling' and (most frequently) 'demonstration' effect [ibid.: 357, 360, 369, 377). A demonstration effect seems the best description of how the diffusion of democratisation globally and regionally can have a cumulative influence. For continual demonstrations of successful democratisation in other countries seem to both reduce the military's; political self-confidence and raise the self-confidence of its civilian opponents (see Table 8.1).
East is an awkward anomaly that requires - and has produced - more in the way of explanation than simply pointing the finger at Islam (Salame, 1994).
Furthermore, the third wave of democratisation swept through the other regions of the world in an almost sequential manner: southern Europe in the mid-1970s, Latin America and Asia in the later 1970s and the 1980s, Eastern Europe in 1989, and Africa in the early 1990s. In the African case the regional wave of democratisation not only removed the surviving African one-party states, but also virtually annihilated what had for decades been the world's largest body of military regimes. In 1989 more than 30 of the 48 countries of (sub-Saharan) Africa were ruled by some form of military regime, but six years later there were
192   Non-Democratic Regimes
only three clear-cut examples of such regime (Wiseman, 1996: 1-2). Many experts on Africa have offered regional explanations for this dramatic democratisation {ibid.: 35-6; Bratton and Van de Walle, 1992) but what particularly needs regional explanation is why the African region lagged behind Latin America and Asia before so quickly making up for lost time.
Regional explanations are also required to deal with anomalies in the way that a particular variety of regime has succumbed to democratisation. The communist regimes show a distinct regional variation in their democratisation, for the three major Asian and sole Latin American examples of communism survived the global democratisation that brought down the other communist regimes (and also the other varieties of dictatorship in their own regions); Moreover, there is a sequential anomaly in the democratisation of the other communist regimes. For while the East European cases unexpectedly collapsed in 1989, both traditionally liberal Yugoslavia and the (for several years previously) liberalising Soviet Union lagged behind and were still not centrally democratised as late as 1991. This anomaly is partly explained by regional explanations of why the six East European communist regimes were so vulnerable to democratisation in 1989.
Regional explanations of the dramatic collapse of the East European regimes in late 1989 have tended to emphasise three factors (Berglund and Dellenbrant, 1991):
• There was a legitimacy 'crisis' arising not only from the regimes' lack of democracy and long history of being virtual client regimes of the Soviet Union (the seemingly puppet rulers of its East European satellite states), but also from the regimes' inability to meet economic aspirations encouraged by communist ideology and by comparisons with Western Europe (Holmes, 1986: 100, 102-3; Przeworski, 1991: 2).
• Secondly, there was the change in attitudes in a now liberalising and partially democratising Soviet Union, especially what appeared to be the abandoning of the long-standing Soviet doctrine of intervening militarily to preserve communist rule in Eastern Europe as in Hungary in 1956 and Czechoslovakia in 1968 (Karl and Schmitter, 1991: 158; Stokes, 1993: 21, 99; Przeworski, 1991: 5).
• Thirdly, a regional snowball or domino effect initiated by the Polish communist regime's willingness
Democratisation 193
(a) in early 1989 to formally negotiate with the pro-democracy opposition,
(b) in April to agree to a series of political concessions that included an almost immediate limited democratisation and promises of full democratisation in a few years time,
(c) in June to implement this agreement by allowing free elections to the agreed quota of seats in parliament, and
(d) in August to allow the installation of a predominantly non-communist government (Welsh, 1994: 386; Przeworski, 1991: 3-4, 55).
In addition to regional explanations there is also a need for supplementary theories or explanations that can account for why the military type of dictatorship - the apparently most powerful type - is actually the most vulnerable to democratisation. A striking anomaly of the 1970s-80s democratisation is that until 1989 many military regimes but virtually no party regimes had succumbed to democratisation. Moreover, the 1989 collapse of communism began with the military-led relinquishing of power in Poland (Przeworski, 1991: 4, 6, 78), where communism had been taken over by the military in 1981 (complete with all the trappings of coup, martial law, military junta and eventual 'civilianisation') and was still headed by a military man in civilian garb, General Jaruzelski.
As the military regime is so vulnerable to democratisation, the most likely place to find an analytical framework for examining the democratising demise of dictatorships is in the analyses of the demise of military regimes. After developing such a framework, the next step is to find a typology of transitions (from military dictatorship to democracy) that can provide the basis for a 'global' set of regional explanations of democratisation. Then attention will return to the other types of dictatorship, the personalist and party, which have been more reluctant than the military to relinquish the power they have seized or misappropriated.
The Military's Relinquishing of Power
The military has shown a much greater tendency than parties or personal rulers to relinquish power. In part this is because it
194   Non-Democratic Resirn.es
Demo era tisation 195
is so difficult for any other organisation or group within society to seize power from the military. But the tendency towards relinquishment is more because the military has often seized power without intending to retain it for the longer term. As Nordlinger pointed out, the military's common lack of com-1 mitment to retaining power on a long-term basis 'helps to account for the low average life span of military regimes', which he calculated to be only about five years (1977: 143, 139).
When Nordlinger sought to distinguish the1 paths through which military regimes were replaced by civilian rule, he included voluntary disengagement (with or without infra-military or civilian pressure) as 'by far' the most common path {ibid.: 141). Furthermore, the other two paths that he identified -countercoups and civilian opposition - seem only variations of the standard form of voluntary disengagement, differing only in form and/or in degree of voluntariness. For the several in1 stances he mentioned of a countercoup's leading to the relinquishing of power by the new military government are also vrltimately cases of voluntary withdrawal by the military; the distinctive feature is one of form in the sense of there being a two-stage process initiated by a countercoup. As for the military being 'forced to relinquish their power by extensive civilian opposition', such opposition is described as 'civilian pressures, demonstrations, strikes, and riots', not armed insurrection, and Nordlinger's two examples also involved military disunity and the refusal of some officers to support their leaders' retention of power (ibid.: 139). In fact he argued that no military regime supported by a united officer corps determined to retain power had ever been overthrown by civilians alone.
The military's relinquishing of power is usually described as its withdrawal or disengagement from power or politics. In the early 1990s Welch concluded that there was still no widely accepted 'paradigm' for the study of military withdrawal/ disengagement even though there had been over 80 cases of military disengagement in the 1940s-80s, of which more than a third were 'through a scheduled, planned withdrawal after holding elections' (Welch, 1992: 324, 334).
However, theorists of military dictatorship have long been aware of the issue of withdrawal/disengagement and have addressed it within their wider discussions of military rule. Long before Nordlinger raised the issue, Finer in 1962 had discussed
11 liJ <
ill
»1
3
s •fe
bo
a
m
. S -5
i- cs
v J3 FT-
fn J2 cn
■a c 3 v *
c« ho
•o
<U •—h
"1 ^ ^
rd u ra bn_ h3
O oj c
C cn
S "0
ts c
c E
O OJ
2 !5 2 O)
a ft
a;
& co ft
° 12
u 3 SO/
ft
T3
a
s
rt
C? CO
02
3 o
Bi Oh — 3
-a ,
a 2
E .
) in : < co :
a „ 3 rt o
.S bo bo „., ■c bo
tS s
33 E
0 "3
T3 ;3
2 -a " * IS 00
y a
"S
d §
: aj _
j -< CO 3 ■£ °o
;z;§2
J3 T)
o C
w bo
XJ bo C3
V ZL
u ^
O 1Q
O.soi
d (1) ,—i
g s
T3 XI
.9 "0
o ~ bo ;
to
Z 2 m
I?
■a §
o o
TJ Q
•ii O
u T3
s °
o
a c3 o -o
00 «
Q u
£ o
u u
^ O rt
tS 73 j>
« CO
O llH
w o 2 R v v
ion o
•3 Gj
oj ^ r-e
O C/3 ft
■-1 o
ho o
^3 *e w „3
O oj
Z Q
T3    - i-i
I -a
3
o
bo C
o ft
. S
to
' 3
E x
c 0 E o
J3 O
nl
3 s
£ P
00
V
cx oj ho
> Ms
'SbSS o ■3 3 2
J3 cj
3 ca
J3 C
ft
3
Is
1- Ih
S 0,3
% o
~ Z
bo ■-a oo
c o)
l— H
o Jo
M O CO OR
1^
ftcn
O i-H ft . 3 ih
3 ho
"° "2 a s ■o z
.u ^—
" 3
3 .
faocS 3  R CO
a j> ft
CN3
-j h 3 2 .Oh
oj R
s 65 :
ft2
bo co
o 2
.3. a;
3 K
3ffi 3
«5 hR
i-,"13 0> 3 m  u u
u     m     - u,
o
ho 3
.3 .13 <ii ~ n « ^ \ 3
-•3 -
cj  3 rt
bo„
.5 %
'B S o S
u O U Q
■°
» O
3 3 o
8 S
3 O
v 3
.£P 0
— 3 (S"
Jj « 00
R -d 2
\, 3
C3
3 H
ft o
33 rU
» 3 Oh .3
m i- f-~
■alts
.3 hoO . ,y R ..h3
.2 is 10 .a
—i "31 CO    n i
a;
nl
-1 «-i
ftS s
pS S2^iSS S
_U i-h
"3
iH oj
Efi 3
^ =5
rt O
. 3 w 3
w 5 o
S 2 u
s ij.j-^ bo bo
3 _
33
.s
196   Non-Democratic Regimes
the 'return to the barracks' (1976: 32, 174-8) and Huntinjrton 1 had explored the issue of retaining/relinquishing power in 1 connection with the veto coups of mass praetorianism (1968-233-7). By the 1980s there were more specialised discussions of the topic of withdraWal/disengagement, such as Clapham and Philip's (1985) analysis of the political dilemmas of military regimes, Sundhaussen's (1985) of military-regime stability1 in Southeast Asia, and Welch's of military disengagement in Africa and Latin America (1987: 20-4). Moreover, thesei writers from the 1960s to the 1980s identified a large number of different factors affecting withdrawal/disengagement, with its relinquishing and transferral of power to civilians (see Tables 8.1 and 8.2).
Many of the factors affecting military withdrawal/disengage-ment have been incorporated into systematic analyses of the issue. By the 1980s, Finer was applying his disposition/opportunity framework (see Chapter 3) to the military's withdrawal from power. Although it was Sundhaussen who had first suggested this application, he later preferred a reasons/preconditions framework for analysing military withdrawal (Sundhaussen, 1985: 272-5). Finer, too, abandoned the term 'opportunity' in favour of 'societal conditions which invite withdrawal', and he also revamped the notion of 'disposition' to include necessary conditions as well as motivations (Finer, 1988: 299-305). His framework was therefore less like a disposition/opportunity analysis than Nordlinger's examination of withdrawal had been, as the latter had focused on 'disengagement motives' and had included the opportunity-like proviso that these motives will motivate a withdrawal 'as soon as an acceptable transition to civilian rule can be arranged' (Nordlinger, 1977: 141).
Adapting the Finer-style motive/means/opportunity calculus of usurpation (see Chapter 3) to fit withdrawal would have several advantages, such as being able to apply the same framework to both the entrance and exit of the military from power. It would not cover exit from power, though, in such an involuntary and therefore 'motiveless' form as expropriation through defeat by foreign or domestic forces or misappropriation by a personalist ruler. Moreover, there appear to be so many difficulties in adapting the motive/means/opportunity framework to fit withdrawal/disengagement that it may be better to approach the issue from the perspective of ceasing to retain power
Devwcratisation 197
TABLE 8.2
The calculus of transferral of power -Factors affecting when/how/to whom
lj, y Motive
Means
Opportunity
ll
; jdßological/national 1 ^interest
1. Protecting national :     interest or rr i   ideological goals
{Welch, 1987; 5
Sundhaussen, 1985) iii 2v Protecting regime's
work in pursuit of
national interest
or ideology
(Sundhaussen, 1985;
Finer, 1988)
Corporate self-interest
1. Protecting military and its corporate interests from retaliation/retribution (Huntington, 1968; Sundhaussen, 1985)
2. Protecting its corporate autonomy under future civilian government (O'Donnell and Schmitter, 1986)
Individual self-interest 1. Protecting governing
and internal-security
officers from
retahation/retribution
(Huntington, 1968;
Sundhaussen, 1985;
O'Donnell and
Schmitter, 1986)
Negative: division over issue of appropriate time to withdraw
(Nordlinger, 1977) Positive: internal consensus/cohesion to withdraw (Sundhaussen, 1985; Finer, 1988)
Civilian successors will not ignore interest/policy preferences of the military (Nordlinger, 1977; Sundhaussen, 1985; Finer, 1988; Welch, 1987) Civilian successors are a potentially stable and peaceful party or party system (Sundhaussen, 1985; Finer, 1988; Welch, 1987)
198   Non-Democratic Regimes
rather than that of withdrawal from power. Instead of looking at the loss of power, this framework would be concerned with the continuing or discontinuing retention of usurped power.
The Calculus of Retention
Any calculus of retention will differ significantly in content -in motives and factors - from the calculus of usurpation, described in earlier chapters. Rustow pointed out long ago that those 'factors that keep a democracy stable may not be the ones that brought it into existence', and the same could be said of dictatorships; he himself noted that military dictatorships 'typically originate in secret plotting and armed revolt but perpetuate themselves by massive publicity and alliances with civilian supporters' (Rustow, 1970: 346, 341). It is not only tliis 'means' aspect which will have changed as the military moves from usurping to retaining power. New individual and corporate self-interested motives are likely to develop as the military enjoy the fruits of office, and any initial commitment to ideological or national-interest motives may evaporate.
Moreover, new motives that actually discourage the retention of power will probably arise. Just as there are motives which inhibit the military from seizing power (see Chapter 3), so there are motives which discourage the military from retaining power (see Table 8.1). Therefore, as with the calculus of usurpation, there is a 'balance sheet' of motives 'for' and 'against'; when motives discouraging retention outweigh those favouring retention, the military is on balance motivated to relinquish power rather than to continue retaining it. Even if on balance there is a motive for retention, this may be too weak to motivate the military to retain power when it faces declines in the means and/or opportunity for retention.
A balance-sheet approach also has to be taken to assessing" the means of retaining power and the opportunity for retention. Analysts of military regimes usually mention only 'negative' factors, which detract from or produce a decline in the military's means/opportunity for retention of power (see Table 8.1). But, as Rustow pointed out, there are also factors that keep a (military) regime stable and that military dictatorships use to perpetuate themselves. These factors which favour or contribute to the retention of power can be presented in terms of means
If,'I* ''
i Vi i'.i'n1 Democratisation 199
1     and opportunity, as being 'positive' means/opportunity factors ; ■ '   that have to be weighed against the negative factors in order i  i      to arrive at a balanced assessment of the strength of a regime's i   means or opportunity. For example, Rustow's reference to I,   ''     military dictatorships' perpetuating themselves through civilian ■'   , '■     alliances can be viewed as a positive 'means' factor, favouring | |;|;     or conuibuting to retention. As for positive 'opportunity' factors, a substantial list of opportunity factors favouring retention I   ,   can be produced by simply reversing well-recognised negative 'i     factors (see Table 8.1). For example, the negative factor of 'negative economic trends' can be reversed to become the ,     positive factor of 'positive economic trends', which may or may not be present in any particular regime's balance of opportunity factors.
!' Therefore when analysing a particular regime, 'calculating'
whether that regime will continue to retain power is a matter I .      of assessing the overall balance of positive and negative motives, means and opportunity (see Figure 8.1). The same 'calculation' is required when analysing why a particular regime I! :      did not continue to retain power and instead relinquished it !      through a democratisation or other form of transferring power, ii     Usually a few obvious changes in motives and/or factors can be readily identified as the likely cause of the shift from retention to relinquishment of power. The means/opportunity factors of the calculus of retention can also be used to help assess whether there will be, or why there was, an expropriation of , :      the regime by domestic forces.
|'        The Calculus of Transferral
I The relinquishing of power, though, needs one further appli-
1 cation of motive/means/opportunity - namely, to the actual transferral of power. For clearly the power must be relinquished to some other organisation, social group or person. A 'calculus
i of transferral' (see Table 8.2) provides a systematic examination of the issue of to whom the relinquished power is being transferred, such as to an allied party or to the people through free elections. Although the military's transferral of power is usually
|        viewed in terms of democratisation, in the 1960s Finer and
ji Huntington were interested in the decades-old Mexican example of the military actually building and transferring power to an
Democratisation 201
official party - and since then there have been other examples of transferral to a party, such as in Cuba, Taiwan and Iraq (Finer, 1976: 180; Huntington, 1968: 239-62). : There is also the associated issue of how the transferral takes place. In the case of transferral of power to the people, for instance, this may be through
(a) elections that have been unilaterally scheduled, organised and supervised by the military; or
(b) a negotiated agreement with civilian representatives, covering the whole process of transition to democracy; or
(c) the military simply abandoning power and leaving it to civilians to sort out how democracy will be established.
Finally, there is the 'tactical' issue of when the relinquishment/ transferral occurs. The military's basic motivation to relinquish power may be too weak to avoid delays and interruptions caused by negative factors in the calculus of transferral. Among these negative factors are the self-interested motive of protection from retribution, and the 'means' problem of the military itself being divided over whether it is the appropriate time to withdraw (see Table 8.2). There are also more positive factors to take into account, though, and in fact the calculus of transferral -with its open-ended questions of when, how and to whom — is in some respects the most complex application of the motive/ means/opportunity framework.
Both the calculus of transferral and the calculus of retention are better suited to analysing a particular case than to categorising, comparing and contrasting a number of different cases. For there are simply too many possible combinations of particular motives and means/opportunity factors; the various combinations cannot be presented as a few general or schematic 'types' that can be used for classification and comparison/ contrast on a global, regional or other basis. Such types can be found, though, in the classic study of transitions from dictatorship to democracy, and its typology of transition covers the transferral as well as relinquishment of power by the military.
202   Non-Democratic Regimes Transitions to Democracy
There emerged in 1986 (O'Donnell, Schmitter and Whitehead) the massive collaborative study, Transitions from Authoritarian Ruli:: Prospects for Democracy, containing more than 20 contributions from a wide range of country, regional and thematic experts. Although these contributors usually referred to 'authoritarian' rather than specifically military regimes, virtually all the Transitions examples were of transitions from military regimes to democracy. Most of the country-studies included were of 1970s-80s (actual or likely) transitions from military rule;, and encompassed a wide variety of cases drawn from southern Europe and Latin America. The Transitions study's 'tentative conclusions' were presented in a long essay by O'Donnell and Schmitter. As they acknowledged (1986: 38), their analysis was heavily influenced by Rustow's (1970) pioneering work on transitions to democracy.
He had categorised the transition to democracy into three phases:
1. prolonged struggle between polarised but evenly matched political forces;
2. the negotiation of a compromise agreement by the political forces' leaders which leads to the institutionalising of democratic procedures; and
3. the habituation of democratic procedures, which gradually enlarges the degree and range of consensus {1970: 352-8, 362-3).
O'Donnell and Schmitter followed Rustow in emphasising the role of compromise agreements — which they referred to as 'pacts' - in transitions to democracy, but they retailored his approach to better fit transitions from military rule. They defined a pact as an explicit compromise agreement that (a) seeks to redefine the political rules of the game and (b) is based on mutual guarantees that protect the vital interests of the pact-makers (1986: 37).
Democratisation 203
Pacted Transitions
A military extrication pact precedes any pacts made between political parties or social forces during the transition from military rule to democracy. The extrication aspect of the pact arises from the officers' desire to 'begin to extricate themselves from direct responsibility for ruling', and includes a growing awareness within the officer corps that the regime will soon - 'in the foreseeable future' - have to seek 'some degree or some form of electoral legitimation' (ibid:. 39, 16). The pacted aspect of the military's extrication arises from the mutual guarantees exchanged with civilians. In an extrication pact the military leaders guarantee civilian representatives that there will be a liberalising restoration of individual rights and of some opportunity for civilians to contest the military regime's policies. In exchange the military are guaranteed that civilians will (a) not seek retribution for repressive 'excesses' committed by military officers, and (b) not immediately or too insistently demand democratisation, let alone resort to disruptive or even violent measures against military rule (ibid.: 40).
The military extrication pact seems to arise in situations similar to those in Rustow's model of negotiated compromise between equally matched, stalemated political forces. But after the pact is made, the balance of power or political strength shifts dramatically in favour of the civilians. The military have initially conceded - through the extrication pact - only an immediate liberalisation of the regime and the prospect of limited, not full, democratisation: 'elections scheduled for an undefined future and, then, for insignificant offices only' (ibid.: 57). But this liberalisation leads to such a post-pact strengthening of the civilian public and weakening of the military that the regime is propelled into a full democratisation. For 'once a government signals that it is lowering the costs [in fear and actual injury] for engaging in collective action and is permitting contestation on issues previously declared oif limits', it soon finds itself facing a 'repoliticized' society - what is termed a resurrection of civil society (ibid.: 48-9).
Although this may well involve the resurgence and/or establishment of political parties, most of the political mobilisation usually comes from social groups and organisations (ibid.: 49, 57). In particular, human-rights activists and organisations
204   Non-Democratic Regimes
stimulate the public's ethical revulsion (often exacerbated by evidence of pervasive corruption) towards the regime's activities. Another particularly significant group is the working class, which has (re-) acquired a capacity for collective action and has been radicalised by its years of political and workplace subordination. Moreover, the various groups and organisations involved in this resurrection of civil society may join together to form what is termed a popular upsurge, in which 'all support each others efforts toward democratization and coalesce into a greater' whole' — which calls itself 'the people' (ibid.: 53-4).
With or without a popular upsurge, the resurrection of civil society propels the military into bringing forward the pact-envisaged elections and extending their scope to produce a full rather than limited democratisation (ibid.: 57). In deciding to transfer power to the people, the now morally discredited military is opting for (relatively orderly) party politics instead of street politics, and it also seems to be calculating that instituting democratisation will enable it to divide and conquer opponents and garner support for its favoured candidate/party from a grateful public (ibid.: 57-8). More importantly, the military has no other realistic means of solving its political problem, for the resurrection of civil society has raised to unrealistic levels the 'perceived costs' of either a return to repression or a (counter) coup by hardline officers - even hardliners are 'likely to hesitate before the prospect of provoking a civil war' (ibid.: 53, 55).
However, although the Transitions conclusions lavished much attention on describing pacted transitions, there seem to be few actual examples of this type of transition among its country studies. The Introduction to its collection of Latin American country-studies admits that the only instances of explicit pacting occurred in Venezuela and Colombia in the late 1950s and, to a partial extent, in Uruguay in the early 1980s — where the agreement 'was barely implemented, if at all', and was 'extremely short-lived' (O'Donnell, 1986: 11-12). Nor did an explicit military-extrication pact occur in any of the transitions included in the southern European country-studies.
Other regions of the world seem just as bereft of military extrication pacts. Explicit pacts did not appear in the Asian 1980s-90s transitions from military rule to democracy: in Pakistan, Bangladesh, Thailand, the Philippines and South Korea.
Democratisation 205
Even the numerous African transitions do not show evidence of pacting: more than two dozen military regimes were democratised in 1990-95 but the 'phenomenon of elite pacting' had 'few echoes in Africa' (Wiseman, 1996: 158). The national conferences on constitutional matters that became quite common in parts of Africa in the early 1990s are only superficially similar to an extrication pact. In reality they differed from pacting in not only their size, procedure and wide range of civilian representatives, but also in their effect (ibid.: 84-94). For they tended either to be the medium for surrender agreements (for orderly expropriations of military regimes) or, in contrast, to be diversionary tactics by rulers who went on to thwart the conference's ambitions.
In fact, apart from Uruguay, the only 1970s-90s example of an explicit miUtary-extrication pact seems to have occurred during the Polish communist regime's democratisation in 1989 (Stokes, 1993: ch. 4). As was noted earlier in the chapter, the military wing of the Communist Party had in December 1981 staged a dramatic military takeover aimed at attacking the country's economic crisis and eliminating Poland's massive independent trade-union movement, Solidarity (Brooker, 1995: 210-16). But in 1988 this by now 'civilianised' regime's party leader, General Jaruzelski, agreed to political negotiations with leaders of Solidarity in order to halt waves of (economically motivated) strikes.
The resulting roundtable negotiations eventually produced accords, announced in April 1989, which seem to be a military-extrication pact but one which contained more concessions than a 'typical' extrication pact. For an almost immediate limited-scope democratisation (with elections in which only 35 per cent of parliamentary seats could be contested by non-communist or non-puppet parties) was to be followed by full democratisation in 1993. However, in typical 'pacted' fashion, the limited-scope elections produced an unexpected situation that the regime could not control (though in parliament rather than the streets) as the regime lost control of its long-standing puppet parties. In August, President Jaruzelski accepted a non-communist prime minister and predominantly non-communist government, bringing to a close his military rule as well as the Polish communist regime.
The pacted transition would not seem so rare if it had been depicted in terms of political balance and compromise
206   Non-Democratic Regimes
Democratisation 207
agreements between the military and their civilian opponents rather than explicit pacts. Later analysts of transitions to democracy recognised that they often involve implicit or tacit negotiations and agreements (Huntington, 1991: 114, 139-40 165-7; Haggard and Kaufman, 1995: 118); and the 'pacted' transition could readily be denned as arising from a situation where military and civilians are quite evenly matched in political strength, engage in explicit or implicit/tacit negotiations and compromise agreements, and see the agreed-upon liberalisation develop unexpectedly into democratisation.
Nevertheless, relatively few transitions from military rule to democracy begin with the military and its civilian opponents quite evenly matched in political strength. O'Donnell and Schmitter recognised that transitions may also begin when the military is in a politically weaker or stronger position than it enjoys at the outset of a pacted transition. They acknowledged that the military may be in such a weak political position that it relinquishes power without securing a pacted exchange of mutual guarantees (O'Donnell and Schmitter, 1986: 39). And they also pointed to situations where, in contrast, the military is in such a strong position that it can actually 'dictate the emerging rules of the game' rather than negotiate a compromise agreement {ibid.: 39). Therefore the Transitions conclusions offer three types of transition from military rule to democracy:
1. the dictated transition by a politically strong military;
2. the pacted transition by a military quite evenly matched with its civilian opponents; and
3. what might be termed the 'abdicated' transition by a politically weak military.
Other Types of Transition - The Abdicated and the Dictated
Although the abdicated and dictated types were only briefly described in the Transitions conclusions, at least there was a more explicit analysis of retention/transferral motives and means/opportunity factors than appeared in the description of the pacted type. The abdicated type of transition involves a relatively straightforward and quite familiar type of military re-linquishment/transferral of power (O'Donnell and Schmitter, 1986: 39, 20-1, 35). It occurs when the military is discredited,
in disarray, or under seemingly irresistible pressure (actual or potential) to relinquish power. In addition to these and other familiar means/opportunity factors, Transitions mentions various familiar aspects of the corporate self-interest motive - only one of which had not been identified by analysts of military withdrawal (see Table 8.1). What seems to distinguish this type of transition from the pacted type is that the military views immediate and full democratisation as an acceptable price to pay to enable it to escape a deteriorating situation. There may even be cases where the transition is actually imposed on the regime by its mobilised civilian opponents. On the other hand, Transitions acknowledged that even in an abdicated transition the process can be complicated by the military's (predictable calculus of transferral) concerns about its corporate autonomy under a civilian government, and about the fate of those officers directly responsible for repression (see Table 8.2).
The (Sub) types of Dictated Transition - the 'Relinquishing' and the 'Pacted-like' The dictated type involves a more complex transition than the abdicated type, perhaps because there is less obvious reason for a politically strong military to relinquish power. In fact there seem to be two different types of dictated transition (see Figure 8.2). In the more straightforward, 'relinquishing' type the military initiates a fully intended relinquishment of public offices and powers through democratic elections - and its only pact-like feature is that the military is sometimes willing to negotiate with civilians over the details of the transition. In the other, 'pacted-like' type the military is initially seeking only some form or degree of electoral legitimacy - not a relinquishment of power - but this leads on to a full democratisation in a somewhat similar (though more controlled) fashion to a pacted transition. The relinquishing type is the more common of the two, and it was seen earlier that there were dozens of cases in the 1940s-80s of the military using elections to stage planned and scheduled relinquishments of power. Yet this type of dictated transition was not explored by the Transitions conclusions, which instead described the rarer and more pacted-like type of dictated transition from military rule.
O'Donnell and Schmitter's example of a military dictated transition, the one initiated by the 1964-85 Brazilian regime, was certainly not a case of planned and scheduled relinquishment
Democratisalion 209
.2 <n
3- <»
C   03 o
--   « E C
w I~ fc \p
O -a ■— to
a' o en '4=
o c n 2
— « o cj
s ° s _
o u 5 =
>■
c a
o JZ •12 en
° 3 E g-a. s o a
° » S -i
_C 03
C  W — _
■I § 8 Sg
i 1; s I
3 E dkS
°" 03 O  5 .£
:= —. .t; ra
0C £ .£ 0) >
'5 c co °>
1?£
t
11
d t c a.
t u>_. ]
— c w) 4-.
J: O 0) C
™ .3 c E
x aj 3 as
c ° c .t
O — to
____i, -r; «
E
- a> a-a
c c o 01 £
-q P
> 03
p +j
-° E '5>
3 9m
■3 S '5
1 > 2
» 5 n ,2 o a.
.2 £
us w
----> 3
- t "a
III
.   id as Qi
-■§2:1*1 S f £§-111
re g p j; c
. C   CQ CO
•2 § K
* -S & S »
—>!B p
of power by means of elections. Instead, it was a case of the military's initially seeking some electoral legitimation through instituting limited democratisation or liberalising its democratic disguise, and ending up with a somewhat unexpected but still controlled relinquishment of power. (In Brazil the 1974 liberalisation of its democratic disguise was not followed by full democratisation until the later 1980s, as will be seen in Chapter 9.) In such cases as the Brazilian, one of the reasons for seeking some electoral legitimacy is that the socioeconomic success of the regime has resulted not only in 'a less active and aggressive opposition', but also in the regime being overconfident about the level of its public support, with the regime hoping to secure not just an electoral majority but a 'comfortable majority' (O'Donnell and Schmitter, 1986: 20).
However, this electoral overconfidence is not so much a motive as an opportunity factor, reducing the regime's opportunity for retaining power. For the overconfidence about its popular support encourages the regime to begin an electoral initiative that unexpectedly leads on to full democratisation. The actual motive for this initiative seems to be the desire for:
1. electoral legitimation of an internal succession to the regime's key office (in the Brazilian case it was one army general succeeding another as the country's President); and:
2. the favourable response of international public opinion to the regime's leaders 'following through on their original claims to be preparing the country for a return to democracy' (ibid.).
Although the authors were doubtless well aware of the dominance of democratic ideology in Latin America (see Chapter 5), they do not mention the likely favourable response of domestic public opinion to an electoral initiative or mention the possibility that the military's claims of democratic intent may be quite genuine. In fact the Transitions conclusions never considered the possibility that authoritarian regimes may have ideological or national-interest motives for relinquishing power. Similarly, the Transitions conclusions did not consider the military regimes' tendency towards comparatively short-lived rule (when compared with other authoritarian regimes), even though this tendency was widely recognised by analysts of military regimes and would
210   Non-Democratic Regimes
be recognised in later analyses of democratic transitions (Huntington, 1991: 117; Haggard and Kaufman, 1995: 11, 13)
The Global and Regional Distribution of Dictated Transitions The dictated type seems to have been the prevalent type of transition in Asia during its 1980s wave of democratisation. The most complex example is Taiwan's long transition to democracy, which occurred against the backdrop of a military regime having gradually transferred power to its official party, the Kuomintang, which in turn was only gradually implementing a dictated transition - the martial-law decree inherited from the military regime was not revoked until 1986 (Cheng, 1989: 489)! This dictated transition also displayed some pact-like features, including a continual informal dialogue that led to formal negotiations between the regime and a civilian opposition which had occasionally used sizeable public demonstrations to support its position. The South Korean transition from military rule was a more straightforward case. It not only lacked the extra complications of a preceding military-to-party transferral but also occurred at a more rapid pace and was accompanied by more open or dramatic use of public demonstrations by the regime's civilian opponents. The dictated type also occurred in Pakistan and Thailand, and the only exception to the rule seems to have been the expropriation of Marcos's military-supportive regime in the Philippines by a combination of some military rebels and masses of 'people power'.
In comparison, the dictated type of transition was less prominent in South America's 1980s wave of democratisation. It appeared in Brazil and perhaps Peru (which was on the borderline with an implicitly pacted transition); but the Uruguayan transition was pacted, the Bolivian was a stop-go combination of pacted and abdicated, and the Argentinian was an abdicated transition. Moreover, Africa's 1990s wave of democratisation saw a huge number of transitions from military rule to democracy, but few examples of the dictated type of transition. African military regimes were 'more reticent about handing power back to civilians' than were their South American counterparts, and what have been termed 'managed transitions' were initiated 'either without great sincerity or in response to popular protest and pressures' (Bratton and Van de Walle, 1994: 481).
Demo era tisation 211
Even such a superficial 'global' survey of the incidence of dictated transitions shows the advantages of using the Transitions dictated/pacted/abdicated typology in global and regional comparisons. Moreover, Transitions could claim to have offered a whole new dimension to retention/transferral analysis of 1 particular cases of transition by raising the issue of the relative strength of groups within the military.
:   Hardliners and Sofiliners
I In a pacted type of transition 'there is no transition whose beginning is not the consequence - direct or indirect - of important divisions within the authoritarian regime itself, prin-
' cipally along the fluctuating cleavage between hard-liners and soft-liners' (O'Donnell and Schmitter, 1986: 19). A transition
i does not begin until the softliner faction, group or body of opinion is sufficiently strong to defeat, politically or militarily, the regime's hardliners {ibid.: 16).
The issue of internal divisions within the dictatorship was also raised by later analysts of transition. Huntington (1991) distinguished between 'standpatters' and liberal or democrat 'reformers' and viewed the relative strengths of the two groups as a crucial factor in determining what type of transition would occur. If reformers are in power, there will be transformation (dictated type); a political balance between them and standpatters will lead to transplacement (pacted type); and if standpatters are in power, there will be replacement (abdicated type or expropriation) (Huntington, 1991: ch. 3). The military's 'cohesion' was also an important issue in Haggard and Kaufman's (1995) analysis of transition. As they pointed out, a decline in the consensus within the military about the desirability of retaining power tends to lead to an attempt to negotiate an exit from power (Haggard and Kaufman, 1995: 12, 102).
However, both these later analyses of transition viewed the presence of personalist rule as being the most decisive internal - 'within regime' or 'within military' - political factor affecting transition. Haggard and Kaufman argued that military 'cohesion is greatest where personalist rulers have gained control over both the government and the military establishment', and that such personalist-based cohesion strengthens the military regime's capacity to withstand pressures to relinqxiish power
212   Non-Democratic Regimes
Democratisation 213
(1995: 12, 79). Huntington viewed personal dictatorships as a separate type of non-democratic regime and argued that 'leaders of personal dictatorships were less likely than those of military and one-party regimes to give up power voluntarily' (1991: 110-1], 120-1). And when Transitions remarks about personalist rulers are examined, it appears that the dictated/pacted/abdicated typology will have to be expanded to accommodate the peculiarities of personalist rule.
Transitions from Personalist Rule
Within Military Regimes
In the Transitions conclusions, O'Donnell and Schmitter pointed out that none of the personalist military dictators mentioned in the country-studies had initiated transitions, which therefore had to wait for the dictator to die or be overthrown by a military countercoup (1986: 34—5). A famous example of transition following removal by death was the manner in which the Spanish transition to democracy took place only after General Franco's long-lasting personalist rule finally ended in 1975 with his death by natural causes. However, the personalist ruler's removal from power by death, incapacity or even retirement does not necessarily
open the way for a transition to democracy. As was noted in Chapter 5, there are several instances of a personalist ruler heading a ruler-type military regime, in which the.military as an organisation is independently committed to the permanence of some form of military rule. In such cases the military can hardly be expected to institute a transition to democracy just because the personalist ruler happens to have died or been incapacitated. For example, after Nasser's death in 1970 the Egyptian military backed the succession of another ex-officer, Sadat, and after his less personalist rule was ended by assassination in 1981, the military supported the succession of Vice-President and ex-General Mubarak - who in turn has established a new form of mildly personalist rule.
As for removal by a military countercoup, this is unlikely when the regime is so highly personalist that the military has lost its professional/corporate autonomy and become an instrument of personal rule. In fact O'Donnell and Schmitter noted that a
civilian 'armed insurrection seems the only way for regime change and eventual democratization' in such 'sultanistic' dictatorships as the Somozas' former regime in Nicaragua (1986: 32-3). However, less highly personalist rulers are much more likely to be removed by the military (whether in a factional or corporate coup), and in fact this is the commonest form of countercoup-produced transition. Three of Nordlinger's four examples of countercoup-produced two-stage withdrawals involved the military removing personalist military dictators in 1955-58: Peron of Argentina, Rojas Pinilla of Colombia and Perez Jimenez of Venezuela (1977: 140). Nor has this been a solely 1950s phenomenon. In 1989 General Stroessner's long-standing personalist regime in Paraguay (see Chapter 6) was overthrown by a coup led by his senior army commander, who then presided over a long transition period that culminated in the 1993 elections. Such anti-personalist democratising countercoups are also to be found outside Latin America, as in the classic example that occurred in the African state of Mali in 1991-92 (Vengroff and Kone, 1995).
The countercoup which removes a personalist ruler is usually motivated by a desire to take the military out of politics in order to prevent any further erosion of its professional integrity (O'Donnell and Schmitter, 1986: 34-5). However, the question of motive is significant only when analysing particular cases, such as when applying the calculus of retention. The more important question is how to categorise these cases of a two-stage relinquishing of power produced by an anti-personalist democratising countercoup.
The Anti-Personalist and Personalist Types of Transition Although it will add further complications to an already complex typology, an 'anti-personalist' type should be added to the dictated/ pacted/abdicated typology of transitions to democracy. After all, the countercoup that removes the personalist dictator is an integral part of two-stage transitions and may have as great an impact upon how the transition unfolds as an explicit pact does in a pacted type of transition. For example, the military's removal of a hated dictator may politically strengthen a formerly discredited and demoralised military, allowing it to carry out a dictated rather than abdicated style of transition. In contrast, the ending of a dictator's repressive rule may lead to a
m
214   Non-Democratic Regimes ■■■.■..a
popular upsurge that leaves the military with little option but to seek an abdicated style of transition.
The Transitions typology of transitions has to be expanded anyway to incorporate a personalist type of transition. For although1 personalist rulers are typically reluctant to relinquish power, there have been occasions when a (not highly) personalist ruler has in fact initiated a transition from military rule. Among the more recent cases were Pinochet's 1988-89 democratising referendum and elections in Chile, and the young military leader of Ghana, Rawlings, completing a transition to democracy in 1992. Therefore, if it is to cover the full range of transitions from military rule to democracy, the Transitions typology needs ' to be expanded to include a 'personalist' as well as an 'anti-personalist' type (see Figure 8.3). In the latter type the removal \ of a personalist ruler begins what will become a dictated, pacted or abdicated style of transition; in the former type a personalist ruler himself begins what will become one of these styles of transition. Some new motives and factors will have to be added to the calculus of retention when it is applied to a case of personalist relinquishment of power, but such factors as the 'electoral option' are better described in relation to personalist transitions that involve a civilian personalist ruler and an official party.
Within Party Regimes
The general tendency of personalist rulers to refrain from relinquishing power is found among party as well as military dictatorships. And in fact by the 1990s there were analyses of transition focused specifically on neopatrimonial personalist types of regime, irrespective of whether the regime was military or civilian.
Snyder's (1992) analysis of the removal of these intransigent neopatrimonial personalist rulers, civilian as well as military, argued that an autonomous military is the most likely remover of a personalist dictator. He contended that only if 'the military lacks sufficient autonomy' to remove a (civilian or military) personalist ruler, is there then an opportunity for him to be removed by a revolutionary movement (Snyder, 1992: 380-1). In their later, regional study of African transitions from (civilian as well as military) neopatrimonial rule, Bratton and Van
liiftii1.
IT
Democratisation 215
FIGURE 8.3
Two additional (personalist) types of transition to democracy
Personalist
Anii-personalist
(Usually nor highly) personalist ruler initiates transition
Dictated style
Pacted style
Abdicated style
(Usually not highly) personalist ruler removed by military countercoup or party
i i
Military or party then 'completes' transition
Dictated style
Pacted style
Abdicated style
De Walle noted the general reluctance of neopatrimonial rulers to relinquish power, and confirmed that this reluctance is most pronounced in the most extreme case of neopatrimonial rule, the 'personal dictatorship' (1994: 462, 475). Therefore, African dictatorships' frequent tendency to degenerate into personalist rule may partly explain why Africa lagged behind other regions in shifting from dictatorship to democracy; its dictatorships' reluctance to relinquish power is what would be expected of personalist rulers (Wiseman, 1996: 18-19; Bratton and Van de Walle, 1994: 454, 459).
Anti-Personalist Transitions Within Party Regimes It was in the Eastern European communist state of Romania that a civilian personalist ruler paid the ultimate price for his reluctance to relinquish power (Stokes, 1993: 52-8, 158-66). By the 1980s the party boss, Ceausescu, had established a quasi-monarchical and repressive form of personalist rule, but his personalist regime quickly crumbled when in December 1989 he prodded the military into a bloody attempt to repress the public protests which had broken out in the city of Timisoara. A mixture of military opposition, internal revolt within the party, and quickly spreading popular uprising led to his execution on 25 December and the party rebels' NSF movement presiding over democratisation.
216   Non-Democratic Regimes
Some six weeks earlier, the much milder party boss of communist Bulgaria, the elderly Zhivkov, had been peacefully removed from power by senior communists in the first stage of an anti-personalist transition (Stokes, 1993: 147-8; Bell, 1993-86-9). The new leaders soon publicly committed themselves to democratisation and in January the regime began roundtable negotiations about elections (held a few months later) with the Union of Democratic Forces, which had been formed by the now many pro-democracy parties and organisations - most of them created after Zhivkov's removal.
Personalist-type Transitions and the Electoral Option In African party dictatorships, though, long-standing personalist rulers reluctant to relinquish power were pressured by the public into a personalist transition. By 1990 only four one-party states were still in the hands of their original and now aged founding leaders: Kaunda in Zambia, Banda in Malawi, Houphouet-Boigny in the Ivory Coast and Nyerere in Tanzania (who was soon to retire in favour of his deputy). In Zambia and Malawi the no-longer-popular founding leaders resisted for a time the growing pressure for democratisation, but eventually held democratic presidential and legislative elections in which they and their decrepit parties were defeated (Van Donge, 1995; Venter, 1995).
In contrast, founding leader Houphouet-Boigny of the Ivory Coast responded to public demands for multiparty elections by quickly initiating a dictated style of personalist transition that gave him a new degree of electoral legitimacy (Crook, 1995); As he was still respected and even admired by much of the public and had revitalised his party in the 1980s, he was in a strong electoral position and in fact went on to win over 80 per cent of the vote in the presidential election, and led his party to an even larger victory in the legislative elections.
Therefore the prospect, or even only the possibility, of winning democratic elections may be a major factor in personalist rulers' decisions to initiate dictated-style transitions or pressured abdicated-style transitions. It must be included as an opportunity factor whenever the calculus of retention is being applied to a personalist ruler. For just as a successful military regime's overconfidence about its popular support can lure it into an electoral initiative, so may a personalist ruler view competing in democratic elections as a more attractive option than con-
"IP]]i|'||l||||l|ll"i|l 1
i11 j'15 [j 11 1 Democratisation 217
|! hjlii11!1,1
i iV   tinuing to struggle to hold on to power dictatorially. Some ' I ill f !   overconfidence about electoral prospects is likely to be found not only among the more successful personalist rulers, but also among those rulers whose backs are against the wall - irrational '   llfljIiJi 1 assessments of election prospects are understandable instances
of wishful thinking in adversity (Przeworski, 1991: 65). ,1 ' lijl'ii       Moreover, the electoral option is also available to military personalist rulers. For example, in the earlier-mentioned case 1        of Rawlings in Ghana he had good prospects for electoral success, ' 1 having provided the country with a decade of 'purposive, effective,
Mi1    dynamic and relatively incorrupt personalist rule', and his elec-i ' V\\i'    toral prospects were likely to decline if he did not soon fulfil [;       his already frequent promises to 'pass power to the people' i      (Haynes, 1995: 99, 101). Similarly, when Pinochet held the (constitutionally prescribed) referendum on whether he should continue as President or hold democratic elections, he doubt-' i'i,;    less felt some hope of winning the referendum - and in fact it:      he did manage to win some 44 per cent of the vote. It is in I        this somewhat paradoxical combination of (a) reluctance to |        relinquish power, and (b) hope of electorally regaining relinquished power that the personalist ruler, whether civilian or jjj,    military, is more like a ruling party than a ruling military.
i      Democratisation of Party Dictatorships
■ v ■
, I      The Durability of Party Dictatorships
i     Until the 1980s party dictatorships had displayed a marked
' reluctance to relinquish power. In contrast to the transitoriness of the common form of military regime, party regimes seemed
■ more committed to retaining power and almost invariably were brought to an end only by armed force. The two fascist dictatorships had been destroyed by foreign invasion in the 1940s, and many of the African one-party states were overthrown by military coups from the 1960s onwards. Five of the newly 1 established African one-party states had been removed in the mid-1960s and the attrition continued at a much reduced rate into the 1970s-80s (McKown and Kauffman, 1973: 56, 56 table 1;
1 Decalo, 1989). But this had been a vulnerability to military coups, not to instituting transitions to democracy. (Only Senegal had
218   Non-Democratic Regimes
Democratisation 219
seen a transition from one-partyism to a form of multiparty democracy and, as will be seen later, there was reason to be dubious about its democratic credentials.) Moreover, until the 1980s none of the world's many communist regimes had succumbed to military coup or democratisation, and in fact they appeared to be the most durable of dictatorships. There were no theories and analyses of party 'withdrawal' or 'disengagement' from power.
Such durability was not simply a result of the party regimes having become personalist and therefore displaying the typical reluctance of personalist rulers to relinquish power. It is true that most party dictatorships have degenerated into personalist rule, but those which escaped this trend (or revived after the death of their personalist ruler) have been no less reluctant to relinquish power. As has often been pointed out, the greater durability of the party regime, as compared to the common form of military regime, is partly due to the party's 'stronger' motivation to retain power, with a party usually differing from the military in its views of ideological rectitude, the national interest and corporate and individual self-interest (as was seen in earlier chapters).
What makes the party dictatorships' durability all the more striking is the party's relatively weak means of retaining power, when compared to the coercive capacity of a military regime. In most cases the party is similar to personalist rulers in having to rely on its control over the military to secure such a massive capacity for the use or threat of force. Therefore the party (and most personalist) dictatorships are vulnerable to expropriation by a military coup, as in the many African cases, and to expropriation by a revolution, as when the communist regimes in East Germany and Czechoslovakia were overthrown in late 1989 by two of history's very rare examples of peaceful and spontaneous revolution.
So it is likely that in some cases a decision to relinquish power was swayed by doubts about whether the military could be relied upon to defend the regime. The military's failure to do so during the collapse of communism in Eastern Europe has been likened to 'the proverbial "dog that did not bark'" (Bunce, 1995: 98). Such political unreliability must also have had a 'demonstration effect' that probably reached as far as Africa. In fact parties' and rulers' doubts about the political
I J, i reliability of their military may have been one of the reasons J1!1   , why such a large proportion of the many party and party-personalist 1 .    dictatorships diat came to an end in the 1980s-90s did so through relinquishments of power.
Considering how many of the party regimes opted to relin-h i quish power, it is not surprising that the Transitions approach (O'Donnell and Schmitter, 1986) was soon applied to analysing transitions from party dictatorship to democracy. The influence of the Transitions approach and terminology is evident in comparative studies of party dictatorships' transition to democracy (a) in Latin America, Southern and Eastern Europe, (b) in Central and Eastern Europe, and (c) in the Soviet Union I (Karl and Schmitter, 1991; Welsh, 1994; Bova, 1991). There i'[ was a tendency, though, to focus on the pacted type of transition rather than applying all three basic types suggested by the Transitions approach. The basic dictated/pacted/abdicated typology, plus the occasional 'surrender' to expropriation, is in fact all that is required when classifying (non-personalist) party dictatorships' transitions to democracy; thankfully there is no need to develop any new types.
Applying the Calculus of Retention and Transferral
■ In contrast, the calculuses of retention and transferral require some significant changes to their lists of motives, means and opportunity when they are applied to party dictatorships. As was noted earlier, the party has different and stronger motivation than the military to retain power but has inherently weaker coercive means of doing so - having to rely on control over the military as its means of wielding military-style coercion. This inherent weakness in the party's coercive means of retaining power cannot be alleviated by the inherent strength of its noncoercive means, such as its possession of politically skilled personnel and an organisation specialising in political mobilisation and indoctrination. Therefore loss of control over the military will always be a crucial 'negative' factor that may nullify a party's means of retaining power and will dominate any party's calculus of retention.
Similarly, the differences between a ruling party's and a ruling military's 'life after dictatorship' have important implications for parties' retention/relinquishment as well as for the 'to whom,
220   Non-Democratic Regimes
Democratisation 221
how, when' of transferral. A party which relinquishes power lacks the secure corporate role and individual careers awaiting an army which returns to the barracks. Moreover, the party will also lack the military's capacity to exercise indirect rule, from the barracks, over at least those policies that affect its corporate interests - as occurred in Brazil and, to varying degrees, in other Latin American countries after their 1980s democratisation (Stepan, 1988: chs 6-8; Pion-Berlin, 1992). Nor will the party usually have the military's at least potential capacity to return to power through a coup or the threat of a coup.
But, unlike the military, a party does have an immediate 'second chance' to regain power constitutionally; like a personalist ruler, it can use democratic means to reacquire public office and powers. In fact a party usiially has better prospects than a personalist ruler of political survival under democracy; the party faces lower 'downside' risks and has better long-term prospects of eventual success. For even if a party loses the initial democratic elections, it may (1) win sufficient seats in parliament to participate in a coalition government or play a significant role in the legislature, and/or (2) survive 'to fight again another day' in later elections (Huntington, 1991: 120).
The Electoral Option
The electoral option can therefore be a particularly powerful factor in a party's decision to relinquish power, especially when accompanied by overconfidence or wishful thinking. It is very likely that the electoral option had a major influence on the earlier-mentioned Romanian and Bulgarian cases of civilian anti-personalist transition. And the Communist Party leaders who, respectively, rebelled against Ceausescu and removed Zhivkov saw their electoral expectations realised when their renamed and revamped parties duly won the first post-dictator elections. So it is not surprising to find ruling parties in won-personalist regimes, too, opting to relinquish power in the expectation or hope of regaining it through democratic means.
However, the Sandinista revolutionary regime in Nicaragua provides a classic example of electoral overconfidence (Vickers, 1990; Williams, 1994). The Sandinistas instituted a dictated transition in the late 1980s with every expectation of electoral
triumph and hoping that an unquestionable democratisation would reassure potential foreign-aid donors, remove the threat of US military intervention and end the war with the counterrevolutionary Contras. In fact the party was so unprepared for defeat in the 1990 elections that a post-election pact had to be negotiated to ensure its smooth exit from office.
In Eastern Europe a similar result occurred when the communist regime in Hungary confidently sought a dictated type of transition (Stokes, 1993: 91, 100-1, 132-4; Swain and Swain, 1993: 69-74). Political reformers began to take over the regime in early 1989 and initially hoped to use a controlled transition to tame the burgeoning pro-democracy opposition and remain the country's most powerful political force. By June the Communist Party was seeking to secure its place in the transition by entering roundtable negotiations with the democratic parties, now united into the Opposition Roundtable. But it took until September to reach an agreement, and further controversy and revisions would see the election date put back until March/ April 1990. The increasingly powerful democratic opposition was able to transform the drawn-out negotiations into a virtual surrender by the communists - leading to Hungary's democratisation being dubbed the 'negotiated revolution'. Moreover, although the communist leaders believed they could perform very well in democratic elections, their democratised and renamed party would actually win less than a tenth of the seats in the 1990 parliamentary elections.
In contrast, the African ruling parties proved remarkably successful in winning transition elections. The remaining handful of party dictatorships which had not fallen to military coups in the 1960s-80s and were free of personalist rule in the early 1990s were able to reacquire electorally their relinquished power. In Tanzania the recent successor of founding leader Nyerere quickly instituted in 1990 a preemptive-procrastinating dictated transition that preempted the development of pro-democracy opposition but inordinately prolonged the transition process, with elections not being held - and won - until 1995 (Baregu, 1994: 169-70). In neighbouring Kenya a much more established successor-leader, Moi, and his party were much more reluctant to begin transition but were still able to win the 1992 elections, thanks largely to the splitting of the opposition vote among three parties (Wiseman, 1996: 60-1, 108, 135-6). A similar
Hip
222   Non-Demo era tic Regimes
approach was taken by the Cameroon ruling party and its well-established successor-leader, Biya, producing similarly unconvincing election victories that again relied largely on a splintering of the opposition vote (Van de Walle, 1994: 143-7).
However, not all the African party dictatorships completed a transition to democracy. In Sierra Leone the process was aborted in 1992 by a military coup, and in Angola it ended in the UNITA opposition claiming electoral fraud and returning to its civil war against the MPLA regime. Despite the Angolan debacle, one of the features of African transitions from dictatorship to democracy was the avoidance of state disintegration in a region that has suffered from ethnic/tribal divisions and even civil wars (Wiseman, 1996: 111; Clapham and Wiseman, 1995: 223-4). Although several African states did collapse into anarchy in the early 1990s, they did not do so while engaged in democratisation. The lack of state disintegration during democ-ratisation was partly due to such 'technical' factors as the absence of federalism but, whatever the reason, the result was a striking contrast with the fate suffered by Yugoslavia and the Soviet Union as their communist regimes collapsed.
Democratisation and Disintegration - Yugoslavia and the* Soviet Union The communist regime in Yugoslavia had decentralised so much power down to the six republics comprising its federal state that democratisation proceeded separately in the various republics, with elections being held in March-April 1990 in Slovenia and Croatia, but not until December 1990 in Serbia, Montenegro, Macedonia and Bosnia. More importantly, democratisation quickly took on an ethnic-nationalist aspect as not only the regional remnants of the Communist Party (notably in Serbia under Milosevic's leadership) but also the democratic opposition played the nationalist card - with Bosnia seeing each of its three ethnic groups establish its own party. The separatist tendency led in 1991-92 to secessions by Slovenia, Croatia, Macedonia and Bosnia - producing a rump Yugoslavia comprising only Serbia and Montenegro.
The huge multiethnic federal state known informally as the Soviet Union (formally as the Union of Soviet Socialist Republics) did not survive the partial democratisation of its communist regime, disintegrating completely in 1991 into its
Democratisation 223
15 federated republics (see Exhibit 8.2). As will be described in the next chapter, not all the new states that emerged from the former Soviet Union would soon complete the transition to democracy - some would remain what will be termed 'protodemocracies'.
Protodemocracies
A protodemocracy arises when an emerging democracy suffers from serious limitations that prevent the transition to full democracy from being completed. These limitations raise concerns about whether democracy will in fact be attained and whether the country might even slide back into a form of dictatorship. A protodemocracy can continue in this state of limbo for several years, extending well beyond the period where the calculus of transferral is applicable, and the situation may involve undemocratic elements that played little or no part in the initial transition from dictatorship to protodemocracy. Therefore the forms of limited democracy associated with protodemocracy are quite different from the explicitly limited democratisation that may occur in the initial stages of transition from dictatorship, such as in the Soviet Union in the late 1980s, in Communist Poland in April-August 1989, and in the plans of a military regime instituting an explicitly pacted transition.
The question or issue of whether a country is protodemocratic rather than truly democratic arises mostly in transitions from party dictatorships, especially from personalist party regimes. The issue is particularly likely to arise when a former ruling party and/or personalist ruler wins the democratic elections which are meant to complete the transition to democracy. An early case of dubious and seemingly incomplete democratisation occurred in the West African Islamic country of Senegal. Years before the 1989-93 wholesale democratisation of party dictatorships, Senegal experienced a dictated, unpressured transition from one-party state to multiparty democracy. The transition had been initiated in the mid-1970s by the country's personalist leader and philosopher-poet, President Senghor, and had apparently been completed soon after his 1980 retirement by his protege-successor, the technocratic President Diouf.
However, the massive electoral victories won by Diouf and
224   Non-Democratic Regimes
III
■II'M I
Exhibit 8.2   Disintegration of a Superpower: The Soviet Union
The most surprising and significant example of democratisation1 of a party dictatorship in the 1980s-90s occurred in the communist superpower, the Soviet Union; but the process was not completed before this federal state disintegrated into its constituent republics (most notably the republic of Russia), thereby ending a superpower as well as the oldest communist, and party, regime (Brown, 1996; Gill, 1994; Bova, 1991). The Union of Soviet Socialist Republics or 'Soviet Union' was a formally federal state comprising fifteen republics (often with a strong ethnic-national aspect) that was under the centralised control of the Communist1 Party of the Soviet Union. In the later 1980s the new party and regime leader, General.Secretary Gorbachev, and his reform coalition instituted a revitalisation or reinvigoration of the Soviet system - including a liberalisation and limited democratisation. Within a few years there were indications of a personalist transition to: democracy as Gorbachev moved to strengthen his personal position (culminating in 1990 with his indirect election to the powerful new state post of executive President), encouraged a weakening of the party's control, and sought to win over the party to a social-democratic rather than Leninist approach to politics.
But the limited democratisation of the late 1980s produced a 'democratic opposition' that became so splintered and diverted by ethnic-nationalist concerns that in 1990-91 the issue of separatism increasingly overshadowed democratisation. The separatist issue took on a constitutional as well as political aspect when the massive Russian republic declared its political sovereignty and the supremacy of its laws over those of the Soviet Union. As most other republics followed suit, by 1991 the Soviet Union was in danger of disintegrating into its component republics.
President Gorbachev was able in April-August 1991 to negotiate a treaty, with Russia and some other republics, which would have secured a continuing but much looser federalism. However, the formal signing of the treaty was preempted by the attempted anti-Gorbachev coup in August, which had the typically radicalising effect of a failed counter-revolutionary coup. Moreover, President Yeltsin of the Russian republic had become a hero as the leader of the widespread opposition to the coup, capitalising on his democratic prestige as Russia's directly elected President. Therefore Gorbachev was in no position to resist the pressure to wind up the Soviet Union. In December it was replaced by a loose confederation termed the Commonwealth of Independent States.
Democral.isa.tion 225
his party in the elections of 1983 and 1988 seemed to fall well short of the democratic ideal. In addition to opponents' claims of electoral malpractices (and to the biased ban on electoral coalitions), there was the more important issue of whether the state's patronage resources and opportunities had been used to reduce the competitiveness of elections. (As will be seen in Chapter 9, this subtle tactic has been used by democratically disguised dictatorships to ensure that their elections are only semi-competitive.) Even a sympathetic analyst of Senegalese democracy emphasised the role of the Senghor-constructed party machine - and the support it received from local Islamic leaders and organisations - in delivering the massive, crucial rural vote to Diouf (Wiseman, 1990: ch. 9). Less sympathetic analysts contended that the Senegalese electoral game was played with 'loaded dice' and pointed to the use of the state's patronage resources to 'buy' the electoral support of these locally influential Islamic figures (Tordoff, 1993: 116-7). Similarly, an account of Senegal's new 'semidemocracy' argued that democracy was limited by the influence of the state and by patron-client politics - and that local Islamic leaders maintained their hold on the countryside (Coulon, 1988).
The post-1988 wave of democratisation of party dictatorships brought many new cases of dubious, incomplete democratisation. For example, in addition to such debatable cases as the Bulgarian ex-Communist Party's election victory or Houphouet-Boigny's overwhelming victory in the Ivory Coast, there were several relatively clear-cut cases of elections falling short of democratic standards (Bell, 1993: 88-9; Crook, 1995: 13-20). In Europe, the Romanian ex-communists' National Salvation Front used undemocratic measures to help win overwhelming presidential and parliamentary victories (Stokes, 1993: 174-5). In Africa, the ex-ruling parties in Kenya and in Cameroon benefited from undemocratic practices as well as from the splitting of the opposition vote when they won their (earlier-cited) unconvincing election victories. These protodemocracies and their counterparts in the former Soviet Union are also examples of the wider notion of 'semidemocracy' which will be described in the following chapter.