4
The Implications of Prior Regime Type for Transition Paths and Consolidation Tasks
aving analyzed the necessary conditions for a consolidated democracy and then spelled out the key differences among the four ideal-typical non-democratic regimes, it should be clear that the characteristics of the previous nondemocratic regime have profound implications for the transition paths available and the tasks different countries face when they begin their struggles to develop consolidated democracies. Within the logic of our ideal types, it is conceivable that a particular authoritarian regime in its late stages might have a robust civil society, a legal culture supportive of constitutionalism and rule of law, a usable state bureaucracy that operates within professional norms, and a reasonably well-institutionalized economic society. For such a polity, the first and only necessary item on the initial democratization agenda would relate to political society—that is, the creation of the autonomy, authority, power, and legitimacy of democratic institutions. We argue in chapter 6 that Spain, in the early 1970s, approximated this position. However, if the starting point were from a totalitarian regime of the communist subtype, democratic consolidation would entail the task of simultaneously crafting not only political society and economic society, but also every single arena of a democracy as well. The full implications of these arguments are spelled out in a more systematic and detailed manner in tables 4.2 and 4.3, but here let us first depict the argument in its most stark form, table 4.1.
The analytic utility of distinguishing between post-totalitarian and totalitarian regimes should now be clear. As table 4.1 demonstrates, it is conceivable that a post-totalitarian regime could begin a transition to democracy with a combination of low-medium or medium scores on each condition necessary for a consolidated democracy except for the autonomy of political society. Hungary in early and mid-1989 came closest to approximating this position. While the tasks facing democrats starting from a mature post-totalitarian regime are challenging, they are substantially less than those facing democrats starting from a totalitarian regime. However, it should also be clear that, precisely because post-totalitarian regimes have a prior totalitarian period, there will be legacies to over-
56
Theoretical Overview
Table 4.1. The Implications of Prior Nondemocratic Regime Type for the Tasks of Democratic Consolidation
Arena Characteristics	Authoritarian	Totalitarian	Post-totalitarian	Sultanistic
Civil society autonomy	Medium to high	Low	Low to medium	Low to medium
Political society autonomy	Low to medium	Low	Low	Low
Constitutionalism and rule of law	Low to high	Low	Medium	Low
Professional norms and autonomy	Low to high	Low	Low to medium	Low
of state bureaucracy				
Economic society with a degree of	Medium to high	Low	Low to	Low to
market autonomy and plurality of		(Communist)	low-medium	medium
ownership forms		or medium (Fascist)		
Note: The character of the arenas in the prior nondemocratic regime in the period relatively close to the start of the transition is of the greatest importance for the tasks democratic leaders will face. The less developed the arena, the greater the tasks democratic leaders will have to accomplish before the new regime can tie a consolidated democracy.
come that are simply not found in an authoritarian regime that has never been totalitarian.
Sharp differences between authoritarian and sultanistic regimes in our typology also help direct attention to the fact that the immediate implications of a sultanistic regime for democracy-crafters (as in Haiti) are that they will have to begin the construction of civil society, constitutionalism and a rule of law, professional norms of the bureaucracy, economic society, and political institutions from a very low base.
The delineation of the different regime types also allows us to be more specific about the possibilities and limits of "pacts" as a transition option available or not available in any particular nondemocratic regime type. Before discussing under what conditions pacts are possible, three general analytic points about pacts must be stressed. First, neither theoretically nor historically do democratic transitions necessarily involve pacts. Indeed, of the eight distinctive paths to redemocrati-zation Stepan analyzed elsewhere, only three involved pacts.1 Second, pacts can range from very democratic to very nondemocratic in their intention and consequences. A pact might be specifically crafted to provide for the rapid dismantling of a nondemocratic regime and the setting of an early and specific date for free elections. Such a pact would be clearly democratic in its intention and, if implemented, its consequences. Or a pact may explicitly entail some nondemocratic constraints for a short period before and after the first foundational election. In contrast, a consociational pact that is not initially undemocratic, if maintained too long, might preclude the entry into politics of new groups and eventually
i. See Alfred Stepan, "Paths toward Redemocratization: Theoretical and Comparative Considerations," in Guillermo O'Donnell, Philippe C. Schmitter, and Laurence Whitehead, eds., Transitions from Authoritarian Rule: Comparative Perspectives (Baltimore: Johns Hopkins University Press, 1986), 64-84,170-74.
^ a>
co co
tu 2
rz -£i o
<d tl 5 '
U    _ 03
I if - •
O   0) t
» i -I
CÜ o
š i
2;s
rape
S S ° S
•é 8 1 g S--B ö
CD  o co "£= o
g g ě s-sž s
o E
o qj F
CD   O   > ni
g S § «> in P. <o M t
Ü > -a -5 a; ■-= o > «3 cj
-= -= "E '5 -1 = £
co
TO c ^ I
o -a cd 3
2 R 5 5 <S St
S. cu -g — —        -t^ m
O- 03  cz -a
id cj I 2
CO CD CD CD
5
CO
CO CD
cd o
d CD'S Q-
QJ c 2 "o   cd cd
3   CD   Q)   C   r 3
m *-=  y  co ^ i±
as — to f
o to
cd a. cz
,C   Q. O
^ o
^ o
o   "£ _cd
co .03 cd
Q3 .5" cd
—    L— O
CD   0   ^ CD
— — CO
-o 0
.§ I
Is
O CL
CD -Í3 O CO
S. 8
to 2
2 "O
to CD.
co a
is {s
■g I «
p clu
tI3 J2
03 0) cz
03 ^
§ "8 s -
o »:|
£r° §.■ ö -§ ts
M I? 03 >§"§■
ÖSE
er o ^ co co ^ > co co
co \z3 o co
, CJ
s s ts
ra.Í b ' ! O. 03 o
i 5" £ E
i -ÍS S "° .
;S a> -o
■S 5 cz
- r>  & CO
i ^5 s
C;   CO CO
•2 Š o S 8. E
S co -o 03
co £
"ti C03 ej co o3
3   - I
li*-
o CO
9- » 5
O c o
03 O s
£ t3 S
^ o> m
O 03 Í=
s.
E -o
a: a.
CO CO
íť   co ~ ^ O
ca o £ E
cl cd
to -a
cd
"o cd
'š ~
■a aS
- J
> o co
E "a c
.« "5 cog:
cö " E
-t—1 cj co
— CD 'iZ
co cl ca
o S £
g E
10   TO co'
? S £
= 3 2?
co ^ co
=3 -1
> ±= -a
.S2 5 S
co
4^ Q.
to
co >-
-Q
II
•c E
co oj
% i,
jCD " öö CD
o o
^ co —
co o E
co E co
cl -.p
§ E
CO o CD
ZJ  g >-«
to tz i_ to
to   CD CO O
co "a tu cl
g. s
0 E
03 CO
.2   SS g
oj ro co
CD O
x E
CD CD
= CO CO ±ť
co
> s
tu co 'í:
B E
O 03 o "o
CO 03 O = .5= E
"no
o £ E .
Sel
CO   .tí    _  tť M-
. "
O)   O) °
>   <D E
5 CD
8. o -a
oj ta E o
co —
g "5     ■ 2 E >
u ■= CD CO
^ E
O CD O "O
'S _^
~   co cd
G  cd cz
™ co
■—
co -S" S
O,   03 CO
fi ü E
Table 4.2. (continued)
Path
3. Interim government after regime termination not initiated by regime (coup by nonhierarchical military, armed insurgents, or mass uprising and regime collapse)
Authoritarianism
In an authoritarian regime, it is possible that an organized democratic opposition in civil society and even political society exists. If they demand early elections, this transition path is quite possible. However, in the absence of effective demand for elections, the interim government will be tempted to exercise revolutionary power in policy areas and to postpone or cancel elections, thus delaying the transition or leading to a new nondemocratic regime.
Totalitarianism
An interim government is unlikely. However, should a deep crisis lead to a successor government, given flattened civil society and the absence of organized democratic political society, successful pressure for the holding of free elections is unlikely. The successors might search for electoral legitimation, but this does not ensure democratization.
Post-Totalitarianism
Early elections are only the most likely path in mature post-totalitarianism where opposition activists might form government and proceed to democratization. In early or frozen post-totalitarianism, the most likely regime transition is mass uprising which, if not repressed, could lead to regime collapse and an interim government. The interim government may well be formed by elites connected with the old regime who are able to consolidate their power electorally in the still "flattened society."
Sultanism
High chance that "interim government" will claim to act in the name of the people and will postpone elections in order to carry out reforms. Given previous lack of autonomy of civil or political society, there is a high chance that groups associated with the sultan but claiming legitimacy for having supported the uprising will achieve nondemocratic power. The best chance for democratic transition is if revolutionary upheaval is led by internationally supported, democratically inclined leaders who set a date for elections and allow free contestation of power.
Table 4.2. (continued)
Path
Authoritarianism
Totalitarianism
Post-Totalitarianism
Sultanism
4. Extrication from rule by hierarchically led military
5. Some regime-specific possible transition paths and likely outcomes
If a regime is led by a hierarchical military, the "military as institution," if it feels under internal or external threat, may play a role in pressuring the "military as government" to withdraw from direct rule and to hold "extrication elections." The length of transition and the extent of the "reserve domains of power" the military can impose as the price of extrication decrease with the severity of the internal or external threat to the military as institution and the strength of democratic forces in civil and political society.
If nondemocratic authoritarian regime is led by nonhierarchical military and this regime collapses or is overthrown, it will be easier to impose civilian democratic control and trials on the military than if the regime had been led by a hierarchical military.
Path not available to this regime type. Primacy of revolutionary party and unconstrained role of leaders make rule by hierarchical military impossible.
Leadership of totalitarian regime could split, opening the way for popular mobilization, liberalization, and possibly even an interim government that holds elections. Given the level of control prior to the mobilization of protest, a more probable outcome is that the dynamic of mobilization leads to re-imposition by force of totalitarian controls or to shift to post-totalitarianism. See transition paths open to post-totalitarianism.
Path not available to this type given leading role of the party.
A post-totalitarian regime, confronted with a serious crisis, could collapse if the option of repression is unavailable. Collapse could lead to non-democratic takeover by alternative elites, democratization, or chaos.
Path not available to this regime type. Sultanism implies a degree of fusion of private and public, and the sultan's interference with bureaucratic norms is incompatible with rule by a hierarchical military.
Given dynastic tendencies of sultanism, if sultan dies of natural causes family members will attempt to continue sultanistic regime; thus, normally no regime-led liberalization will take place.
03 ^_
"S ~  zj ~
5^
00 ±3
1 g
CZ O
f    § 8
E =■
co co
co
o -a
43   CJ   =2 CD
CD
CD £
a. cu ■
g -=
cn c:
CD   CO £3
o a.
2 * 03 eg aj °
O   03   03   O   X Q.
_c _rz Q a. as o
^ TO  £, =>
p _   r-n 03 r-n O s
CD CO
^ 03 2 e S € £ £
O Q_ O
=3 O _
g ™ g
_ co -->
■r; ^ & CZ    CD O
ti   (D   3 U
8 .52
CD "O
^ CD
O CD
O CD
H a- °
• ^ E
-w  g .S2
J§ ° c g
"5   °"  03 O
b; ^ co
03 2 E „
_CZ   t aj CO
5- c T « o 6
co   cd o   cd cd c
- 1 «
cd co o
-a cd
cd -ci
£ "> s °-s- =
o _q" q3 co
S5  > ~
■ 5SSE
+=• -= CD
1 = ° ~ g °
E
CD -t-J
p .,      — CD
5> E
co £
d CD CO "O
o o -je
F P c S ?  S  to Y
O   O- CD   CD   1— +-»
r—>   Q_ CD CD
co 00
13 O
CD   O   CD .
u x: -a ^ ^ 5
2
00 "o
CD J±!
«    0   ■ =
> _ o
O — CD
P   X 5
CO -O
■11
CD o
* ° o
-a a. .2
£ s- ™
> V) Q)
<o S » 5 o j c .9- . = S 0 » S iS w03ocD-=ca5-S
5j ™ | -5) I - I E
g .52
03 CD
!    CL CO
f=   CD CD
cd-q. E
CD
CD
-Q CD -i-< Q-
C C a)
0 o > E E -
CD CD
01 CD CD CD —
O CD
03 _g-2
03   O   CD O
P > m 2.
CD
.O 2=  Q. :
wj i_   ^   CD ■
O u i; r >
CL >   CZ CD
CL O   O m-
o cn cj o —
o F j_
" § o
cd "e H~
E .5 «
'a?B S E
^ -— O. CO
!= S 0 E
CD O c CO
™ O 'i-
SS t/3 S s
— O CO ^
co a. cz co
o <0 S
* E
- 03 s ffi
CO t! CD CD
J2 "D  co co
' 03 !^ 1 -C   03 o
1   ~ .1 CO
CD X J= j= -a ^3
cz s 5 <» o
ca ■= -
03 .CO CD cj it
6   03 .
^ cti
CD
3 <->
o » .
Q. 03 I
CD 03
■g -2 co
CD O
CZ O -o
CO E   03 £-
CZ CZ   CD —
O O CD
o c i: E
^5"
Transition Paths and Consolidation Tasks
61
become a form of "exclusionary consociational authoritarianism."2 Finally, a pact that is designed to exclude some groups permanently and vastly to over-represent other groups is clearly undemocratic in its intention and, as long as it is sustained, in its consequences. Third, as Stepan has argued elsewhere, "pact creation does not necessarily mean pact maintenance—pacts can fall apart____Pacts—with or without consociational elements—cannot be created in all political systems. Party pacts have two requirements: first, leaders with the organizational and ideological capacity to negotiate a grand coalition among themselves; second, the allegiance of their political followers to the terms of the pact."3
Much of the transition literature on pacts contains references to "hard-liners" and "moderates." Transitions are frequendy seen as involving a pact between the regime moderates and the opposition moderates who are both able to "use" and "contain" their respective hard-liners. This is, in essence, a four-player game theory model.4 However, two conditions must be satisfied for it to be a true four-player game. The moderate players in the regime must have sufficient autonomy so that they can, over time, conduct strategic as well as tactical negotiations with the players from the moderate opposition. Conversely, the moderates in the opposition need a degree of continued organizational presence, power, and followers in the polity to play their part in the negotiation pacts. For many writers on transitions, the locus classicus of such a pact occurred in Spain.5 In Spain, as we shall see in chapter 6, regime and opposition moderates initially crafted a pacted reform. Eventually, negotiations led to a pacted rupture that allowed the dismantling of the nondemocratic elements of the Franco state and the creation of new democratic structures. This overall process is called reforma pactada-ruptura pactada.
While there are often references to the possibility of pacts being a key part of most transitions, full four-player pacts are possible only in two of our four ideal-typical nondemocratic regimes. A regime that approximates the sultanistic ideal type does not have the reforma pactada-ruptura pactada available as a transition path because the two moderate players are absent. The essence of the sultanistic ideal type is that the sultan fuses personal and public power. Important figures in the regime are significant not because of any bureaucratic or professional position they hold, but because of their presence on the personal staff of the sultan.
2. Jonathan Hartlyn discusses consociational exclusion in The Politics of Coalition Rule in Colombia (Cambridge: Cambridge University Press, 1988).
3. Stepan, "Paths toward Redemocratization," 80. Stress in original. For an excellent analysis of the difficulties of pact maintenance, see Eric Nordlinger, Conflict Regulation and Divided Societies (Cambridge: Center for International Affairs, Harvard University Press, 1972).
4. See, for example, Adam Przeworski, "The Games of Transition," in Scott Mainwaring, Guillermo O'Donnell, and Samuel Valenzuela, eds., Issues in Democratic Consolidation (Notre Dame, Indiana: University of Notre Dame Press, 1992), 105-53.
5. For a rigorous and appropriate application of the game theory approach to the Spanish case, see Josep M. Colomer, Game Theory and the Transition to Democracy: The Spanish Model (Aldershot, England: Edward Elgar Publishing, 1995), and "Transitions by Agreement: Modeling the Spanish Way," American Political Science Review (December 1991): 1283—1302.
A
Table 4.3. The Implications of Nondemocratic Regime Type for the Minimal Tasks of Completing Transition to and Consolidation of a Democratic Regime from that Regime Type
Necessary Conditions
Authoritarianism
Totalitarianism
Post-totalitarianism
Sultanism
1. Rule of law and civil society freedom
2. Political society autonomy and trust and legal condition for it
In some authoritarian regimes there is a tradition of rule of law and civil society that might be quite lively, but civil liberties will need to be extended and protected. Laws giving autonomy to trade unions, media, etc., may need to be enacted and implemented.
All the normal conditions ensuring the free electoral competition between parties need to be created. In some cases, party competition has only been suspended and can easily be revitalized. In other cases, the formation of parties needs to be legalized and restrictions on specific parties lifted. In some cases the political rights of key political actors need to be re-established. In exceptional cases an authoritarian state party may have to be dismantled.
Rule of law did not exist. Much of the legal code, to the extent that it existed, was highly politicized and instrumental for the party-state but not for its citizens and therefore was incompatible with democracy. Civil liberties are minimal and need to be legalized, developed, and protected. The "flattened" nature of civil society requires fundamental changes that are difficult to generate in a short time.
The party's dominant position in all areas of society and its privileged status and resources must be dismantled, its presence in all institutions removed, and almost all of its property transferred to the state. However, if citizens want to recreate the party they should be allowed to do so, and its support and power should depend on the votes people might want to give to it. Given the flattened social landscape the representation of interests will be particularly difficult.
An extensive reform of the legal system to assure civil rights and rule of law will be needed.
The dismantling of the privileged status, legal and otherwise, of the dominant party will be needed. Legal reform will also be needed to assure the free formation and competition of political parties. While society may not be as "flattened" as under totalitarianism, the relative lack of economic and political differentiation makes political "representation" of interests difficult and complicates the development of a normal spectrum of democratic parties.
Given the legacy of the fusion of public and private and the extreme personalization of power, the establishment of a rule of law and guarantees for citizens have a high priority and will be a difficult task.
The suppression of semiprivate violence and the creation of a modicum of trust are requirements for the development of political parties, free contestation for power, and sufficient autonomy for the working of democratic procedures and institutions.
Table 4.3. (continued)
Necessary Conditions
Authoritarianism
Totalitarianism
Post-totalitarianism
Sultanism
3. Constitutional rules to allocate power democratically
4. State bureaucracy acceptable and serviceable to democratic government
In some cases, there can be an immediate declaration that a previous democratic constitution has been reinstated; in other cases amendments to a nondemocratic constitution may be viable; in still others a full democratic constituent assembly and constitution-making process are needed.
To the extent that the bureaucracy has not been politicized and has maintained professional standards, there may be no immediate need for bureaucratic reform. In some cases, a more or less limited purge of bureaucrats, including the judiciary and the military, might be desirable. But if a hierarchical military played a major role in the previous nondemocratic regime, such purges may be quite difficult.
A paper constitution may exist that, when filled with democratic content, might lead to perverse consequences, since it was not designed for a democratic society. The making of a new democratic constitution will be necessary but difficult due to an inchoate political society, the lack of a constitutional culture, and the legacy created by the verbal commitments of the previous constitution.
The delegation of major tasks of the state to the party and the penetration of the party into all bureaucratic and social institutions make the creation of a nonpoliticized bureaucracy an imperative and difficult task. The dismantling of the party within the state might seriously reduce the efficiency and coordination of the state apparatus and open the door for a clientelistic take-over by the new democrats or by opportunists. The experience of the party state leaves a legacy of popular distrust of the state.
Given the fictive character of the constitution, there are serious costs to using these institutions, and the making of a democratic constitution should be a high priority.
The fact that many functions of the state, including judiciary functions, were performed by party bureaucrats makes purges and reform of the state bureaucracy a widespread demand but a complex and contentious issue to resolve. The skills of the former bureaucratic elite and the lack of experience of the opposition may well give the former elite a privileged position.
A universalistic legal culture will have to be developed. Even while there may be a usable constitution, given the recent abuse of constitutional rules, a spirit of trust and respect for constitutionalism does not exist at the end of a sultanistic period.
The clientelistic penetration and corruption of bureaucratic institutions limit their efficiency and legitimacy and put extensive reform on the agenda. Even democratically elected leaders may perpetuate clientelistic practices rather than rational administration.
co ^ _S
o _
= co ~
E as 5
>=   3 CO
ca o 03
°- ^ =
Qj  -t^ CD
1 s
0 :-t±
03 Q.
€ _- .&
> > .32
1 " 8
= m— co
CO O
) CD
^ E c
CO   n O
CO   o O
CO -— 03
CD CD _
CJ   > "O
._, CD
2 E
£ o
: CD O
' .Izl CD
! 8 8
: CL c _
i "O w a
' cz eo
j CD Q) CJ
O CD
CD CZ CO o
e   cd   > — £
_     "r 1    ^> CO
§ 12 •
o E co c:
o ° o) a- -a 8 E 2 .32 co
=   CD   03   O o
DconTlca^lru
e
- o o < E
^ >-"2
.32 E o
K ° "~
o cz
" o c
r. U Q]
.— 03 CZ
§ 1 if
E ~
$2 €
5 3 « f
E -§
•6 2
0)
E   w Q3
CO w
E •& 1 °
I J3 "(D
: Z3 »-
I O -
■ +9
■ CD EZ
; q. g
i co
cd -a
tz co
CD OJ
a. CD
E
+± co
a? 5
CD CO
E
2 E
I I S3
13 E « 03 fo
1.2 —
g | & 03 ^ CO u-
o u_    cd °
O c cz O co cz   cd   o   cd co
±±  o  w 'z;       cd ■
5 c
o —
CD CD
CD
^ 5=   CD CD
\Z3 CO CO   CO   CD CD
.E E 5 E
88»
u   ™ CO
I ^ °
*«— CO
Si? W
CD :
CD >j=
CD  +3 X3 oj o
■ — cd ^ x: .p
-n   Z3 CD ■—
~ ^ CD o
I-    X CO CD
E B
co cd -
i—    > _ -
03 £
£2 o 2r
to "n e
92 o co E
E E
03 = >- 03 S   ID Oil
e e-g-s
03 co O   Q- _ C
<_ g co .2
f- .7=
^ g E
CD CD
JZ! O u-
■♦—1 CZ CD
co    co '
CO CD
■g €
03   e w
■a .P = X £
co    n o
■^1 O 03
03 CO 03 o
E
£ M "
" e
E
>- o I °
= — £r
CO   S 03
|i'8
P- o <•» 03   Q. o
4-. CO CD
C >- r- —
.9 E ,
o o A
(3 8 ;
Transition Paths and Consolidation Tasks
65
But there is absolutely no room on the "household" staff of the sultan for a moderate player who publicly negotiates the demise of his employer. The other players who never exist in an ideal typical sultanistic regime are moderates from the organized democratic opposition. Neither civil society nor political society has enough autonomy to enable a publicly organized democratic opposition to develop sufficient negotiating capacity for it to be a full player in any pacted transition.6
A similar logic would preclude the ideal-typical totalitarian regime from even a full two-player game. There is a big player (the totalitarian hard-line maximum leader and his party-state staff) and a small underground opposition (half a player?) that can struggle to exist and possibly resist but that has absolutely no capacity to negotiate a pacted transition.
Even early post-totalitarian regimes do not have sufficient diversity and autonomy in the ruling party-state leadership or sufficient strength and autonomy within the democratic opposition really to produce all the players needed to conclude successfully a four-player democratic transition game. Indeed, as we argued in chapter 3, if an early or a "frozen" post-totalitarian regime faces a crisis of opposition, it is particularly vulnerable to collapse if it is not able to repress that opposition, given its limited negotiating capability. But a mature post-totalitarian regime (such as Hungary in the mid-1980s) and a wide range of authoritarian regimes (such as Spain and Brazil in the mid-1970s) can produce four-player games. Thus although "pacted transitions" figure prominently in the literature, the classic four-player pacted transitions are in fact available as a transition path only in some authoritarian and mature post-totalitarian regimes.
A transition path that would seem available to most nondemocratic regimes but that, upon closer scrutiny, is in fact available only to the authoritarian regime type concerns the military. If the costs of rule by the "military as government" are considered too great for the "military as institution," a free election may become part of the extrication formula for the hierarchical military in charge of an authoritarian regime.7 However, the control of the government by a hierarchical military bureaucracy is completely inconsistent with the logics of sultanism or totalitarianism or of the leading role of the party in post-totalitarianism.
We are now ready to present for analysis a resume of the implications of non-democratic regime types for paths to democratic transition (table 4.2), and of the implications of nondemocratic regime type for the minimal tasks of completing the transition to and consolidation of a democratic regime (table 4.3).
6. For example, the only Warsaw Pact country in 1988 not to have one opposition samizdat journal published in the country was Romania, a country that combined under Ceausescu strong sultanistic and totalitarian tendencies. For the special difficulties of a successful democratic transition from a sultanistic regime, see the introductory chapter by H. E. Chehabi and Juan J. Linz in their edited volume in progress, Sultanistic Regimes, and Richard Snyder, "Explaining Transitions from Neopatrimonial Dictatorships," Comparative Politics 24 (July 1992): 379-99. Also see Michael Bratton and Nicholas van de Walle, "Neopatrimonial Regimes and Political Transition in Africa," World Politics (July 1994): 453-89.
7. An extensive conceptual and political analysis of the distinction between the "military as government" and the "military as institution" is developed in chapter 5.