© 1996 The Johns Hopkins University Press

All rights reserved. Published 1996

Printed in the United States o f America on acid-free paper
05 04 03 02 01 00 99 98 97 96 5 4 3 2 1

The Johns Hopkins University Press

2715 North Charles Street
Baltimore, Maryland 21218-4319
The Johns Hopkins Press Ltd., London

Library o f Congress Cataloging-in-Publication Data

will be found at the end o f this book.
A catalog record for this book is available from the British Library.

ISBN 0-8018-5157-2 ISBN 0-8018-5158-0 (pbk.)

 1

Democracy and Its Arenas

I n p a r t l, we explore how the character of different nondemocratic regimes

affects, or does not affect, the paths that can be taken to complete a transition to
a democratic regime. We also explore what implications prior nondemocratic
regime types have for the probable tasks that must be undertaken before such
fledgling democratic regimes could be considered consolidated.
Having structured our question thus, our argument cries out for definitions of
completed democratic transition and consolidated democracy. In this book, when­
ever we attempt to establish how far any given country has gone toward completing
a transition to democracy, the definitional standard we will use is the following:

A democratic transition is complete when sufficient agreement has been reached about politi­
cal procedures to produce an elected government, when a government comes to power that is
the direct result of a free and popular vote, when this government de facto has the authority to
generate new policies, and when the executive, legislative and judicial power generated by the
new democracy does not have to share power with other bodies de jure.

With this working definition, it should be clear why democratic activists and
theorists insist on distinguishing between liberalization and democratization. In
a nondemocratic setting, liberalization may entail a mix of policy and social
changes, such as less censorship of the media, somewhat greater space for the
organization of autonomous working-class activities, the introduction of some
legal safeguards for individuals such as habeas corpus, the releasing of most polit­
ical prisoners, the return of exiles, perhaps measures for improving the distribu­
tion of income, and most important, the toleration of opposition.
Democratization entails liberalization but is a wider and more specifically po­
litical concept. Democratization requires open contestation over the right to win
control of the government, and this in turn requires free competitive elections,
the results of which determine who governs. Using these definitions, it is obvious
that there can be liberalization without democratization.1

l. Much o f the conceptual confusion about what Gorbachev was and was not doing with glasnost and
perestroika could have been avoided by a clearer understanding o f these two concepts. The above discus­
sion o f the difference between liberalization and democratization is drawn from Alfred Stepan, Rethinking
Military Politics: Brazil and the Southern Cone (Princeton: Princeton University Press, 1988), chap. 1. Also
see Guillermo O’ Donnell and Philippe C. Schmitter, Transitions from Authoritarian Rule: Tentative Con-
4 Theoretical Overview

Our discussion of what constitutes a completed democratic transition helps

highlight two further issues. First, transitions may begin that are never completed,
even though a new authoritarian regime does not assume power. Our definition
helps guard against the “electoralist fallacy,” that is, that a necessary condition of
democracy, free elections, is seen as a sufficient condition of democracy.2 Some
of the most common examples of electoralist nontransitions are found in those
cases where a previously ruling military (e.g., in Guatemala in the 1980s), though
relinquishing direct control of government, retains such extensive prerogatives
that the democratically elected government is not even de jure sovereign. Second,
by including in our definition the need to reach an agreement on the specific in­
stitutional arrangement for producing democratic government, we are alerted to
decision-making within the democratic political arena. Disagreements among
democrats over such issues as a unitary versus a federal state, a monarchical or re­
publican form of government, or the type of electoral system may create ques­
tions about the legitimacy of the emerging democratic government, the decision­
making process, and indeed the future of the political system. Such institutional
indeterminacy about core procedures necessary for producing democracy may
not only leave the transition incomplete, but also postpone any consolidation of
democracy. We do not mean that there cannot be disagreement about the most
desirable democratic institutions early in the transition and in the constitution­
making process. Such disagreement is normal. But a deep and continuous con­
frontation and ambivalence about democratic institutions among the political
elites and the majority of the population, with no sign of accommodation to the
enacted institutions, is certainly not conducive to consolidation. It is, therefore,
disagreement not only about the value of democracy but also about the specific
institutions of a democracy that might make consolidation difficult.
There is a further political and intellectual advantage to being clear about what
is required before a transition can be considered complete. Nondemocratic power
holders frequently argue that certain liberalizing changes they have introduced
are sufficient in themselves for democracy. Introducing a clear standard of what
is actually necessary for a completed transition makes it easier for the democratic

elusions about Uncertain Democracies (Baltimore: Johns Hopkins University Press, 1986), 7-11, Adam Prze-
worski, Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America
(New York: Cambridge University Press, 1991), chap. 2, Guiseppe di Palma, To Craft Democracies: An Essay
on Democratic Transitions (Berkeley: University o f California Press, 1990), esp. 81-89, and Samuel P. Hunt­
ington, The Third Wave: Democratization in the Late Twentieth Century (Norman: University o f Oklahoma
Press, 1991),9.
2. ‘Electoralism,” or what we call the “electoralist fallacy,” figured prominently in recent debates about
Central America. See Terry Karl, “ Imposing Consent? Electoralism vs. Democratization in El Salvador,” in
Paul W. Drake and Eduardo Silva, eds., Elections and Democratization in Latin Americay 1980-1983 (San
Diego: Center for Iberian and Latin American Studies, University of California, San Diego, 1986), 9-36. See
also Abraham F. Lowenthal, ed., Exporting Democracy: The United States and Latin America (Baltimore:
Johns Hopkins University Press, 1991) and Thomas Carothers, In the Name of Democracy: U.S. Policy toward
Latin America in the Reagan Years (Berkeley: University o f California Press, 1991).
 Democracy and Its Arenas 5

opposition to point out (to their national and international allies as well as to the
nondemocratic regime) what additional, if any, indispensable changes remain to
be done. Such a standard is also extremely useful in circumstances where the old
nondemocratic regime has collapsed or been overthrown and an interim govern­
ment is in power. Such moments are normally replete with elation, sweeping re­
forms, and decrees. However, unless there is a rapid commitment to completing
all the steps required for a democratic transition, the “temporary” interim gov­
ernment may become permanent.3
In most cases after a democratic transition has been completed, there are still
many tasks that need to be accomplished, conditions that must be established,
and attitudes and habits that must be cultivated before democracy could be con­
sidered consolidated. What then are the characteristics of a consolidated democ­
racy? Many scholars, in advancing definitions of consolidated democracy, enu­
merate all the regime characteristics that would improve the overall quality of
democracy. We favor, instead, a narrower definition of democratic consolidation,
but one that nonetheless combines behavioral, attitudinal, and constitutional di­
mensions. Essentially, we mean by a consolidated democracy a political situation
in which, in a phrase, democracy has become “the only game in town.”4
Behaviorally, democracy becomes the only game in town when no significant
political groups seriously attempt to overthrow the democratic regime or secede
from the state. When this situation obtains, the behavior of the newly elected gov­
ernment that has emerged from the democratic transition is no longer dominated
by the problem of how to avoid democratic breakdown. Attitudinally, democracy
becomes the only game in town when, even in the face of severe political and eco­
nomic crises, the overwhelming majority of the people believe that any further
political change must emerge from within the parameters of democratic formulas.
Constitutionally, democracy becomes the only game in town when all the actors
in the polity become habituated to the fact that political conflict will be resolved
according to the established norms and that violations of these norms are likely
to be both ineffective and costly. In short, with consolidation, democracy be­
comes routinized and deeply internalized in social, institutional, and even psy­
chological life, as well as in calculations for achieving success.
Our working definition of a consolidated democracy then follows:

3. We discuss interim governments in more detail in chapter 5. For an extensive discussion o f such gov­
ernments, seeYossi Shain and Juan J. Linz, Between States: Interim Governments and Democratic Transitions
(New York: Cambridge University Press, 1995), esp. 28-40.
4. For other discussions about the concept o f democratic consolidation, see Scott Mainwaring, Guillermo
O’Donnell, and J. Samuel Valenzuela, eds., Issues in Democratic Consolidation: The New South American
Democracies in Comparative Perspective (Notre Dame, Ind.: University o f Notre Dame Press, 1992). For an
especially rigorous discussion o f the concept, see in that volume J. Samuel Valenzuela, “ Democratic Con­
solidation in Post-Transitional Settings: Notion, Process, and Facilitating Conditions,” 57-104. Also see di
Palma, To Craft DemocracieSy 137-55. We owe the telling expression “only game in town” to Guiseppe di Palma.
6 Theoretical Overview

— Behaviorally, a democratic regime in a territory is consolidated when no sig­

nificant national, social, economic, political, or institutional actors spend signifi­
cant resources attempting to achieve their objectives by creating a nondemocratic
regime or turning to violence or foreign intervention to secede from the state.
— Attitudinally, a democratic regime is consolidated when a strong majority
of public opinion holds the belief that democratic procedures and institutions are
the most appropriate way to govern collective life in a society such as theirs and
when the support for antisystem alternatives is quite small or more or less isolated
from the pro-democratic forces.
— Constitutionally, a democratic regime is consolidated when governmental
and nongovernmental forces alike, throughout the territory of the state, become
subjected to, and habituated to, the resolution of conflict within the specific laws,
procedures, and institutions sanctioned by the new democratic process.
Two important caveats. First, when we say a regime is a consolidated democ­
racy, we do not preclude the possibility that at some future time it could break
down. But we do mean to assert that such a breakdown would not be related to
weaknesses or problems specific to the historic process of democratic consolida­
tion per se, but to a new dynamic in which the democratic regime cannot solve a
set of problems, a nondemocratic alternative gains significant supporters, and
former democratic regime loyalists begin to behave in a constitutionally disloyal
or semiloyal manner.5
Our second caveat is that we obviously do not want to imply that there is only
one type of consolidated democracy. An exciting new area of research is precisely
on the variety of consolidated democracies. We also do not want to imply that
consolidated democracies could not continue to improve their quality by raising
the minimal economic plateau upon which all citizens stand and by deepening
political and social participation in the life of the country. Within the category of
consolidated democracies there is a continuum from low to high quality democ­
racies; an urgent political and intellectual task is to think about how to improve
the quality of most consolidated democracies. Our goal in this book is related but
distinct. Since we are living in a period in which an unprecedented number of
countries have completed democratic transitions and are attempting to consoli­
date democracies, it is politically and conceptually important that the specific
tasks of crafting democratic consolidation be understood. Unfortunately, too

5- In essence this means that the literature on democratic breakdown, such as that found in John J. Linz
and Alfred Stepan, eds., The Breakdown o f Democratic Regimes (Baltimore: Johns Hopkins University Press,
*978), would be much more directly relevant to analyzing such a phenomenon than this current book or
other books on democratic transition and consolidation. We obviously do not mean this as a criticism o f
the transition literature. Rather, our point is that the democratic transition and democratic breakdown lit­
eratures need to be integrated into the overall literature on modern democratic theory. From the perspec­
tive of such an integrated theory, the “ breakdown o f a consolidated democracy” is not an oxymoron.
 Democracy and Its Arenas 7

much of the popular and ideological discussion of the current “wave” of democ­
ratization is dominated by electoralism per se and/or the assumed democratizing
potential of market mechanisms per se. But democratic consolidation requires
much more than elections and markets.

T he F ive A r e n a s of a C o nsolidated D emocracy

We believe that consolidated democracies need to have in place five interact­

ing arenas to reinforce one another in order for such consolidation to exist. There
is an additional factor involved.
Democracy is a form o f governance of a state. Thus, no modern polity can be­
come democratically consolidated unless it is first a state. Therefore, the inexis­
tence of a state or such an intense lack of identification with the state that large
groups of individuals in the territory want to join a different state or create an in­
dependent state raises fundamental and often unsolvable problems. Because such
“stateness” problems are so basic, and so underanalyzed, we devote the next chap­
ter to examining this topic. For our argument here, however, it is enough to say
that, without the existence of a state, there cannot be a consolidated modern dem­
ocratic regime.
If a functioning state exists, five other interconnected and mutually reinforc­
ing conditions must also exist or be crafted for a democracy to be consolidated.
First, the conditions must exist for the development of a free and lively civil soci­
ety. Second, there must be a relatively autonomous and valued political society.
Third, there must be a rule of law to ensure legal guarantees for citizens’ freedoms
and independent associational life. Fourth, there must be a state bureaucracy that
is usable by the new democratic government. Fifth, there must be an institution­
alized economic society. Let us explicate what is involved in crafting this inter­
related set of arenas.
By civil society we refer to that arena of the polity where self-organizing groups,
movements, and individuals, relatively autonomous from the state, attempt to
articulate values, create associations and solidarities, and advance their interests.
Civil society can include manifold social movements (women’s groups, neigh­
borhood associations, religious groupings, and intellectual organizations) and
civic associations from all social strata (such as trade unions, entrepreneurial
groups, journalists, or lawyers). The idea of civil society, as a normative aspiration
and as a style of organization, had great capacity to mobilize the opposition to the
military-led bureaucratic-authoritarian regimes in South America, most notably
in Brazil, and was crucial in Eastern Europe as a vehicle for asserting the auton­
omy of those who wanted to act “as if they were free,” especially in Poland.
In addition to the whole range of organizations, such as illegal or alegal trade
unions, religious communities, bar associations, associations of students and pro­
8 Theoretical Overview

lessors, which constitute the complex web of civil society, we should not forget
another part of society: ordinary citizens who are not a part of any organization.
Such citizens are often of critical importance in shifting the regime/opposition
balance because they turn up in the streets in protest marches, heckle the police
and the authorities, express their opposition first to specific measures, support
broader demands, and ultimately challenge the regime. Normally they are initially
small in numbers and later more numerous and can, in some cases, overwhelm
the representatives of the regime, forcing them to consider a growing liberaliza­
tion and ultimately a regime change. However important, numerous, and heroic
such relatively unorganized groups may be, they would not be able to overthrow
the regime and establish a democratic regime if there were not the processes we
focus upon in this book. The fact that none of the regimes included in our work,
in the transition periods we analyze, was ready to use massive force, to give orders
to shoot on the crowds and thus provoke a massacre, as in Tiananmen Square, has
led us to give relatively little attention to the possibility that these nondemocratic
regimes could have been maintained by force. There is evidence that some lead­
ers considered that possibility. They occasionally alerted their security forces for
combat readiness (e.g., in Berlin the day after the wall was breached). Ultimately,
however, such repression did not happen. The cost of that scale of repression was
too high, and the belief in the legitimacy of such a response too weak. Nondemo­
cratic regimes, at least in southern Europe, the southern cone of South America,
and most of Communist Europe, did not enjoy that kind of legitimacy, and many
ordinary, unorganized people in civil society often discovered it, almost before the
rulers themselves. The most dramatic area where this was so was in parts of Com­
munist Europe, such as East Germany and Czechoslovakia.
By political society in a democratizing setting we mean that arena in which the
polity specifically arranges itself to contest the legitimate right to exercise control
over public power and the state apparatus. At best, civil society can destroy a non­
democratic regime. However, a full democratic transition, and especially demo­
cratic consolidation, must involve political society. The composition and consol­
idation of a democratic polity must entail serious thought and action concerning
the development of a normatively positive appreciation of those core institutions
of a democratic political society— political parties, elections, electoral rules,
political leadership, interparty alliances, and legislatures— by which society con­
stitutes itself politically to select and monitor democratic government.6
For modern democratic theory, especially for questions about how to consol­
idate democracy, it is important to stress not only the distinctiveness of civil soci­
ety and political society, but also their complementarity This complementarity is

6. For an earlier discussion o f the need to distinguish between the civil society arena and the political
society arena, see Stepan, Rethinking Military Politics, chap. 1. For an adroit use o f the concepts in a concrete
historical analysis, see Paolo Farneti, “ Social Conflict, Parliamentary Fragmentation, Institutional Shift,
and the Rise of Facism: Italy,” in Linz and Stepan, eds., The Breakdown of Democratic Regimes, 3-33.
 Democracy and Its Arenas 9

not always recognized. As we document throughout the book, one of these two
dimensions is frequently neglected in favor of the other. Worse, within the demo­
cratic community, champions of either civil or political society all too often adopt
a discourse and a set of practices that are implicitly inimical to the normal devel­
opment of the other. Since this opposition is seldom explicit, let us discuss the
forms this implicit opposition may take and how and why such discourse and
practice impede democratic consolidation.
In the recent struggles against the nondemocratic regimes of Eastern Europe
and Latin America, a discourse was constructed that emphasized “civil society
versus the state.” This dichotomy, of course, has a long philosophical genealogy.7
More importantly for our purposes, this philosophical tradition was politically
useful to those democratic movements that emerged in recent contexts where ex­
plicitly political organizations were forbidden or extremely weak. A conception of
a civil society in opposition to the state was also politically useful as the opposi­
tion attempted to isolate the nondemocratic regime and its state by creating an
“us” versus “them” ethical politics. Civil society in many countries was rightly
considered the celebrity of democratic resistance and transition.
The problem arises at the moment of democratic transition. Quite often dem­
ocratic leaders of political society argue that civil society, having played its his­
toric role, should be demobilized so as to allow for the development of normal
democratic politics.8 Such an argument is bad democratic theory and bad demo­
cratic politics. A robust civil society, with the capacity to generate political alter­
natives and to monitor government and state can help transitions get started, help
resist reversals, help push transitions to their completion, help consolidate, and
help deepen democracy. At all stages of the democratization process, therefore, a
lively and independent civil society is invaluable.
But we should also consider how to recognize conceptually, and thus help
overcome, the false contradictions some set up between civil society and political
society. The danger that democratic groups primarily located in civil society

7. For the contemporary revival o f civil society (especially in Eastern Europe and South America), see
the chapter by that name in Jean Cohen and Andrew Arato, Civil Society and Political Theory (Cambridge:
M IT Press, 1992), 29-82. This invaluable book adroitly combines political philosophy with comparative po­
litical analysis. Also see the introduction to John Keane, ed., Civil Society and the State: New European Per­
spectives (London: Verso, 1988), 1-31. The Cohen-Arato book is also valuable because it does not subsume
political society into civil society. János Kis, an important political philosopher, dissident, and former pres­
ident o f the Free Democrats o f Hungary, was one o f the first thinker-activists to discuss the distinctive arenas
o f civil and political society. See Kis, Politics in Hungary: For a Democratic Alternative (Boulder, Colo.: Social
Science Monographs, 1989).
8. In some transitions the apparent crisis o f civil society merely reflects the fact that some o f the move­
ments called civil society had, to a great extent, been created and directed by previously illegal political par­
ties. When these political parties were able to participate legally in politics, they shifted their efforts away
from the mobilization o f civil society and, in some cases, consciously demobilized society. This phenome­
non was particularly noticeable in Spain concerning those civil society organizations controlled by the
Communist Party. Many civil society movements in Chile were analogously controlled by Christian Dem­
ocratic or Socialist Party leaders.
IO Theoretical Overview

might occasionally present for the development of political society is that norma­
tive preferences and styles of organization, perfectly appropriate to civil society,
might be taken to be desirable or, indeed, the only legitimate style of organization
for political society. For example, many civil society leaders view with moral
antipathy “ internal conflict” and “division” within the democratic forces. Insti­
tutional routinization, intermediaries, and compromise within politics are often
spoken of pejoratively.9 But each of the above terms refers to an indispensable
practice of political society in a consolidated democracy. Democratic consolida­
tion requires parties, one of whose primary tasks is precisely to aggregate and rep­
resent differences between democrats. Consolidation requires that habituation to
the norms and procedures of democratic conflict regulation be developed. A high
degree of institutional routinization is a key part of such a process. Intermediation
between the state and civil society and the structuring of compromise are likewise
legitimate and necessary tasks of political society. In short, political society,
informed, pressured, and periodically renewed by civil society, must somehow
achieve a workable agreement on the myriad ways in which democratic power
will be crafted and exercised.
To achieve a consolidated democracy, the necessary degree of autonomy and
independence of civil and political society must further be embedded in and sup­
ported by the rule of law, our third arena. All significant actors— especially the
democratic government and the state— must respect and uphold the rule of law.
For the types of civil society and political society we have just described, a rule of
law embodied in a spirit of constitutionalism is an indispensible condition. A
spirit of constitutionalism requires more than rule by majoritarianism. It entails
a relatively strong consensus over the constitution and especially a commitment
to “self-binding” procedures of governance that require exceptional majorities to
change. It also requires a clear hierarchy of laws, interpreted by an independent
judicial system and supported by a strong legal culture in civil society.10
The above three conditions— a lively and independent civil society, a political
society with sufficient autonomy and a working consensus about procedures of
governance, and constitutionalism and a rule of law— are virtually definitional
prerequisites of a consolidated democracy. However, these conditions are much
more likely to be satisfied if a bureaucracy usable by democratic leaders and an
institutionalized economic society exist.
Democracy is a form of governance of life in a polis in which citizens have

9 - We discuss at length the question o f antipolitics in our chapter on Poland and in our discussion o f
Czechoslovakia. For an example o f how the language o f “ethical civil society” and “normal political society”
can at times emerge as normative opposites, see our chapter on Poland and especially table 15.1. Antipolitics
also created problems in post-transition Brazil.
to. For an excellent volume that discusses the relationships between constitutionalism, democracy, legal
culture, and “self-bindingness,” see Jon Elster and Rune Slagstad, eds., Constitutionalism and Democracy
(Cambridge: Cambridge University Press, 1988), esp. 1-18.
 Democracy and Its Arenas li

rights that are guaranteed and protected. To protect the rights of its citizens and
to deliver the other basic services that citizens demand, a democratic government
needs to be able to exercise effectively its claim to the monopoly of the legitimate
use of force in the territory. Even if the state had no other functions than these, it
would have to tax compulsorily in order to pay for police, judges, and basic ser­
vices. Modern democracy, therefore, needs the effective capacity to command,
regulate, and extract. For this it needs a functioning state and a state bureaucracy
considered usable by the new democratic government. As we shall see in chap­
ter 2, there are many reasons why in many territories of the world no such state
exists. In this book the question of state disintegration is particularly important
in parts of the former Soviet Union. But, insufficient state taxing capacity or a
weak normative and bureaucratic presence in much of the territory, such that cit­
izens cannot effectively demand that their rights be respected or receive any basic
entitlements, is a great problem in many countries in Latin America, such as
Brazil. The question of the usability of the state bureaucracy by the new demo­
cratic regime also emerges in countries where the outgoing, nondemocratic re­
gime has given tenure (as in Chile) to many key members of the state bureaucracy
carrying out politically sensitive functions in justice and education. Important
questions about the usability of the state bureaucracy by new democrats in­
evitably emerge in cases (as in much of post-Communist Europe) where the dis­
tinction between the party and the state had been virtually obliterated and the
party went out of power, disintegrated, or was deligitimized.
The final supportive condition for a consolidated democracy concerns the
economy, or rather an arena we believe should be called economic society. We use
the phrase “economic society” to call attention to two claims that we believe are
theoretically and empirically sound. First, there has never been and there cannot
be a non-wartime consolidated democracy in a command economy. Second, there
has never been and almost certainly there never will be a modern consolidated
democracy in a pure market economy. If both of these claims are demonstrated to
be sound, modern consolidated democracies require a set of socio-politically
crafted and sociopolitically accepted norms, institutions, and regulations, which
we call economic society that mediates between state and market.
Empirically, no evidence has ever been adduced to indicate that a polity that
would meet our definition of a consolidated democracy has ever existed in a com­
mand economy. But the question persists. Is there a theoretical reason to explain
such a universal empirical outcome? We think so. On theoretical grounds, our as­
sumption is that at least a nontrivial degree of market autonomy and ownership
diversity in the economy is necessary to produce the independence and liveliness
of civil society so that it can make its contribution to a democracy. Likewise, if all
property is in the hands of the state and all price, labor, supply, and distributional
decisions are the exclusive purview of the state in control of the command econ­
 Theoretical Overview

omy, the relative autonomy of political society required in a consolidated democ­

racy could not exist.1 1
But why do completely free markets not coexist with modern consolidated de­
mocracies? Empirically, serious studies of modern polities again and again verify
the existence of significant degrees of market intervention and state ownership in
all consolidated democracies.12 Theoretically, there are at least three reasons why
this should be so. First, notwithstanding the ideologically extreme but surpris­
ingly prevalent and influential neoliberal claims about the self-sufficiency of the
market, pure market economies could neither come into being nor be maintained
without a degree of state regulation. Markets require corporation laws; the regu­
lation of stock markets; regulated standards for weight, measurement, and ingre­
dients; and the protection of property, both public and private. All of these re­
quire a role for the state in the economy. Second, even the best of markets have
market failures that must be corrected if the market is to function well.13 No less
an advocate of the “ invisible hand” of the market than Adam Smith acknowledged
that the state is necessary to perform certain functions. In fact, in a neglected but
important passage in The Wealth of Nations, Adam Smith assigned three indis­
pensable tasks to the modern state.

First, the duty of protecting the society from the violence and invasion of other independent
societies; secondly, the duty of protecting, as far as possible, every member of the society from
the injustice or oppression of every other member of it, or the duty of establishing an exact ad­
ministration of justice; and, thirdly, the duty of erecting and maintaining certain public works
and certain public institutions which it can never be for the interest of any individual, or small
number of individuals, to erect and maintain; because the profit could never repay the expense
to any individual or small number of individuals, though it may frequently do much more than
repay it to a great society.14

The third reason for market intervention in consolidated democracies is that

democracy entails free public contestation concerning governmental priorities
and policies. If a democracy never produced policies that generated government-
mandated public goods in the areas of education, health, and transportation,

n. Robert A. Dahl in a similar argument talks about two arrows o f causation that produce this result.
Concerning the political arrow o f causation (e.g., in the case of Leninism in power), the party-state’s ide-
ology goes explicitly against the autonomy of civil society and political society. Or, in the case o f a state con­
trolled command economy, the economic arrow o f causation goes against certain material needs o f a con­
solidated democracy, such as freedom o f the press, because paper and printing materials could be denied.
See Dahl, “Why All Democratic Countries Have Mixed Economies,” in John Chapman and Ian Shapiro,
eds., Democratic Community (New York: New York University Press, 1993), 35:259-82.
12. See, for example, the ample documentation concerning fourteen advanced democracies contained
in John R. Freeman, Democracies and Market: The Politics o f Mixed Economies (Ithaca, N. Y : Cornell Uni­
versity Press, 1989).
13- For an excellent analysis of inevitable market failures, see Peter Murrell, “ Can Neoclassical Economics
Underpin the Reform of Centrally Planned Economies?” Journal of Economic Perspectives'), no. 4 (1991): 59-76.
14- Adam Smith, The Wealth of Nations (London: J. M. Dent and Sons, Everyman’s Library Edition,
1910), 2:180-81.
 Democracy and Its Arenas D

some safety net for its citizens hurt by major market swings, and some alleviation
of gross inequality, democracy would not be sustainable.15 Theoretically, of
course, it would be antidemocratic to rule such public policies off the agenda of
legitimate public contestation. Thus, even in the extreme hypothetical case of a
democracy that began with a pure market economy, the very working of a mod­
ern democracy (and a modern advanced capitalist economy) would lead to the
transformation of that pure market economy into a mixed economy, or a set of
norms, regulations, policies, and institutions we have called an economic society.16
Any way we analyze the problem, democratic consolidation requires the insti­
tutionalization of a socially and politically regulated market. This requires an eco­
nomic society, which in turn requires an effective state. Even such a neoliberal
goal as narrowing the scope o f public ownership (privatization) in an orderly and
legal way is almost certainly carried out more effectively by a stronger state (in
terms of capacity) than by a weaker state. A severe breakdown of the existing levels
of the economy because of state incapacity to carry out any regulatory functions
greatly compounds the problems of economic reform and of democratization.17
In summary, a modern consolidated democracy can be conceived of as being
composed of five major inter-relating arenas, each of which, to function properly,
has its own primary organizing principle. Properly understood, democracy is
more than a regime; it is an interacting system. No single arena in such a system
can function properly without some support from one, or often all, of the other

15. The working of a modern democracy will normally also result in the legal exclusion o f some activi­
ties from the market, such as the sale o f human organs and the sale o f children for sexual services.
16. Robert Dahl’s line of reasoning follows a similar development. “ Democracy would almost certainly
lead to the destruction of certain economic orders [including] not only a capitalist command economy but
also a strictly free market econom y.. . . If people who are harmed by the market have the freedom, power,
and opportunity to do so they will attempt to regulate the market so as to eliminate, or at least limit, the
damage they perceive . . . Political competition provides elected leaders with incentives for responding to
the views and votes of any organized or unorganized aggregate o f people.. . . One way or the other, then,
over time the victims of free markets are likely to influence the government— or some government, whether
local, state, provincial, or regional— to adopt interventionist policies intended to mitigate the harm— The
upshot is, then, that every democratic country has rejected the practice, if not always the ideology, of un­
regulated competitive markets. Although it is true that a market economy exists in all democratic coun­
tries, it is also true that what exists in every democratic country is a market economy modified by govern­
ment intervention.” See his “Why All Democratic Countries Have Mixed Economies,” 259-82.
17. In post-Communist Europe, the Czech Republic and Hungary are well on the way to becoming in­
stitutionalized economic societies. In sharp contrast, in the Ukraine and Russia the writ o f the state does
not extend far enough for us to speak o f an economic society. The consequences of the lack of an economic
society are manifest everywhere. For example Russia, with a population fifteen times larger than Hungary’s
and with vastly more raw materials, received only 3.6 billion dollars of direct foreign investment in 1992-93,
whereas Hungary received 9 billion dollars o f direct foreign investment in the same two years. Much of the
explanation for this variance was that Hungary had a strong economic society in the area o f a law, contracts,
and regulatory regime, whereas Russia had virtually no economic society and its command economy had
ceased to exist. The direct foreign investment figures are from the Wall Street Journal publication, Central
European Economic Review (Summer 1994), 6. The most cited example of market success, the Czech Re­
public, is in fact one of the clearest examples of the political crafting of a new economic society. Indeed,
Stephen Holmes argues that “ Prime Minister Václav Klaus, far from being the antistatist he pretends to be,
is the most talented state-builder of postcommunist Europe.” See Holmes, “ The Politics of Economics in
the Czech Republic,” East European Constitutional Review 4, no. 2 (1995)- 52 - 55>quote from p. 52.
Table 1.1. The Five Major Arenas of a Modern Consolidated Democracy: Inter-related Principles and Mediating Fields

Prim ary O rganizing N e c e ssa ry Support Prim ary M ediation

A rena Principle from other A re n a s upon other A re n a s

Civil Freedom of Rule of law which establishes legal guarantees Interests and values of civil society are the major generators of
society association and State apparatus to enforce rights of civil society to organize if political society
communication these rights are violated Civil society generates ideas and helps monitor the state
Economic society w ith sufficient pluralism to support the apparatus and econom ic society
necessary degree of autonomy and liveliness of civil society

Political Free and inclusive N eeds legitim acy in eyes of civil society Crafts constitution and major law s
society electoral contestation Needs legal guarantees anchored in rule of law and m aintained M anages state apparatus
by impartial state apparatus Produces overall regulatory fram ework for econom ic society

Rule Constitutionalism A legal culture w ith strong roots in civil society and respected Establishes a hierarchy of norms that make actions by, and
of law by political society and the state apparatus upon, other arenas legitim ate and predictable

State Rational-legal Norm ative support from civil society for rational-legal authority Imperative enforcem ent on civil, political, and econom ic societies
apparatus bureaucratic norms and its attendant monopoly of legitim ate force of dem ocratically sanctioned law s and procedures established by
Monetary support levied by political society and produced and political society
rendered to the state by a functioning econom ic society, which
has produced a sufficient taxable surplus

Economic Institutionalized Legal and regulatory framework produced by political society, Produces the indispensable surplus to allo w the state to carry
society market respected by civil society, and enforced by the state apparatus out its collective good functions and provides a m aterial base
for the pluralism and autonomy of civil and political societies
 Democracy and Its Arenas 15

arenas. For example, civil society in a democracy needs the support of a rule of
law that guarantees the right of association and needs the support of a state ap­
paratus that will effectively impose legal sanctions on those who would attempt
to use illegal means to stop groups from exercising their democratic right to or­
ganize. Furthermore, each arena in the democratic system has an effect on other
arenas. For example, political society crafts the constitution and major laws,
manages the state apparatus, and produces the overall regulatory framework for
economic society. In a consolidated democracy, therefore, there are constant
mediations between the arenas, each of which is correctly in the “field” of forces
emanating from the other arenas (table 1.1).
 5

Actors and Contexts

I n a d d i t i o n to our “macrovariables” of prior regime type and stateness,

we call attention to some other important variables that affect democratic transi­
tion and consolidation and that lend themselves to middle range propositions.
Two actor-centered variables concern the leadership base of the prior nondemo-
cratic regime and the question of who initiates and who controls the transition.
Three context variables relate to international influences, the political economy of
legitimacy and coercion, and constitution-making environments.

T he In s t it u t io n a l C o m p o s it io n a n d L e a d e r s h ip

of t h e P r e c e d in g N o n d e m o c r a t ic R e g im e

Our central question here concerns the core group that is in day-by-day con­
trol of the state apparatus. What is the institutional character of this state elite?
Does its character favorably or unfavorably affect democratic transition and con­
solidation? The organizational base is necessarily analytically distinct from the
variable of regime type because, within some regime types (especially authoritar­
ian), there can be dramatically different types of state elites, each with quite dif­
ferent implications for democratic transition and consolidation. Without being
exhaustive, four different types of state elites can be distinguished: (1) a hierar­
chical military, (2) a nonhierarchical military, (3) a civilian elite, and (4) the dis­
tinctive category of sultanistic elites.

Hierarchical Military
As shown in chapter 4 on the consequences of prior nondemocratic regime
types, only an authoritarian regime has the possibility of being controlled by a hi­
erarchical military organization. Control by such an organization is against the
logics of a totalitarian, post-totalitarian, or sultanistic regime.1 All hierarchical

1. In some cases, such as Chile and Uruguay, and especially the “dirty war” in Argentina, the military de­
veloped a definition of the enemy in their national security doctrine that gave to the repression a totalitar­
ian dimension. See, for example, Alexandra Barahona de Brito, “Truth or Amnesty—Human Rights and
Democratization in Latin America: Uruguay and Chile” (Ph.D. diss., University of Oxford, 1993). 28-61.
 Actors and Contexts 67

military regimes share one characteristic that is potentially favorable to demo­

cratic transition. The officer corps, taken as a whole, sees itself as a permanent
part of the state apparatus, with enduring interests and permanent functions that
transcend the interests o f the government of the day. This means that there is al­
ways the possibility that the hierarchical leaders of the military-as-institution will
come to the decision that the costs of direct involvement in nondemocratic rule
are greater than the costs of extrication. Thus, the reassertion of hierarchical au­
thority in the name of the military-as-institution is a permanent danger faced by
the military-as-government. Furthermore, as members of a situational elite who
derive their power and status from the existence of a functioning state apparatus,
the military-as-institution have an interest in a stable state, and this requires a
government.2 This often means that, if a democratic regime is an available ruling
formula in the polity, the military may decide to solve their internal organiza­
tional problems and their need for a government by devolving the exercise of gov­
ernment to civilians. Paradoxically but predictably, democratic elections are thus
often part o f the extrication strategy of military institutions that feel threatened
by their prominent role in nondemocratic regimes.
We can make parsimonious and much less optimistic statements about hierar­
chical military regimes in relation to democratic consolidation. Precisely because
the military (short of their elimination by foreign powers or by revolution) is a
permanent part of the state apparatus and as such has privileged access to coer­
cive resources, members of the military will be an integral part of the machinery
that the new democratic government has to manage. Theoretically and practi­
cally, therefore, the more the military hierarchy directly manages the state and
their own organization on a day-by-day basis before the transition, the more
salient the issue of the successful democratic management of the military will be
to the task of democratic consolidation. Furthermore, the more hierarchically led
the military, the less they are forced to extricate themselves from a nondemocratic
regime due to internal contradictions, and the weaker the coalition that is forcing
them from office, the more the military will be in a position to negotiate their
withdrawal on terms where they retain nondemocratic prerogatives or impose
very confining conditions on the political processes that lead to democratic con­
solidation. More than any of the three other kinds of organizational bases found
in nondemocratic regimes, a hierarchical military possesses the greatest ability to
impose “reserve domains” on the newly elected government, and this by defini­
tion precludes democratic consolidation. This is a particularly acute problem if

2. For a more discursive argument about the analytical and historical utility o f the distinction between
military-as-government and military-as-institution, see Stepan, “ Paths toward Redemocratization, 75-78,
172-73. For the concept o f the military as a “situational elite” with a special relationship to the state, see Al­
fred Stepan, “ Inclusionary and Exclusionary Military Responses to Radicalism with Special Attention to
Peru,” in Seweryn Bialer, ed., Radicalism in the Contemporary Age (Boulder: Westview Press, 1977), 3:221-39,
344-50.
68 Theoretical Overview

the hierarchical military have been involved in widespread human rights viola­
tions and condition their loyalty, as a part of the state apparatus, upon not being
punished by the new democratic government. Such a legacy of human rights vi­
olations presented severe problems for democratic consolidation in Argentina
and Chile.
This is not meant to imply a static situation. Power is always and everywhere
relational. We simply mean that, if a relatively unified, hierarchically led military
has just left the direct exercise of rule, the complex dialectical tasks of democratic
power creation and the reduction of the domains of nondemocratic prerogatives
of the military must become two of the most important tasks for new democratic
leaders.

Nonhierarchical Military
A nonhierarchical, military-led nondemocratic regime, on the other hand,
has some characteristics that make it less of a potential obstacle to democratic
transition and especially democratic consolidation. Concerning democratic
transition, if a nonhierarchically led military-as-government (e.g., of colonels
and majors) enters into difficulties, the incentive for the military-as-institution
to re-establish hierarchy by supporting an extrication coup is even higher than it
would be if the military-as-government were hierarchically led. The fundamen­
tal political and theoretical distinction, however, concerns democratic consoli­
dation. The chances that the military-as-institution will tolerate punishment
and trials of members of the outgoing nondemocratic government are signifi­
cantly greater if the group being punished is not seen to be the military institu­
tion itself, but a group within the military which has violated hierarchical norms.
Likewise, if the colonels have established para-state intelligence operations that
are perceived as threats even to the organizational military, the hierarchical mil­
itary is much more likely to acquiesce (or even insist) that their reserve domains
of power be eliminated.

Civilian Leadership
In comparative terms, civilian-led regimes (even mature post-totalitarian
civilian-led regimes in which Communist parties are essential components) will
characteristically have greater institutional, symbolic, and absorptive capacities
than either military or sultanistic leaders to initiate, direct, and manage a demo­
cratic transition. Civilian leaders are often more motivated to initiate and more
capable of negotiating a complicated reform pact than are the military. They often
have more links to society than do military or para-military sultanistic leaders.
Civilians also can see themselves as potential winners and rulers in a future dem­
ocratic regime. This option is much less likely for military or sultanistic rulers.
There are, of course, potential problems for full democratic transition and
 Actors and Contexts 69

consolidation in such civilian-led political change. Civilian-led liberalization may

re-equilibrate the system short of democratic transition or allow groups to win
elections by skillful but nondemocratic means because of their privileged access
to levers of power. When we consider democratic consolidation, however, it seems
to us that the capacity of civilian leaders in a previously nondemocratic regime to
create obstacles to democratic consolidation, such as constitutionally sanctioned
reserve domains of power, is significantly less than that of a military organization.
An exception to the above assertion might seem to be the case of a civilian-led,
nondemocratic regime based on a monopoly party— especially a ruling Commu­
nist Party. Should this kind of organizational base be considered an obstacle to
democratic consolidation comparable to a hierarchical military organization that
has just left power? Some political activists in Eastern Europe feared that a de­
feated ruling Communist Party and a defeated ruling hierarchical military were
functional equivalents in terms of their ability to impede the consolidation of de­
mocracy. However, we believe that, in those cases where the Communist Party has
been defeated in free and competitive elections (as in Hungary in 1990), this anal­
ogy is fundamentally misleading on two grounds: (1) organizational relationships
to the state apparatus and (2) incentives. The hierarchical military, unless it has
been militarily defeated and dissolved by the new democratic incumbents, will, as
an organization, withdraw as a unit into the state apparatus where it still has
extensive state missions and state-allocated resources (as in Chile in 1989). A de­
feated Communist Party, in contrast, while it may well retain control of many re­
sources and loyalties that help it compete in later elections, has no comparable in­
stitutional base in the state apparatus, has no continuing claim on new state
resources, and has no continuing state mission. Organizationally, it is a defeated
party out of office and, though it may win open elections in the future (as in Hun­
gary in 1994), it has less collective resources to impose “ reserve domains” than do
the military out of office. Our argument here is restricted only to those cases
where the democratic opposition wins open and contested elections and then as­
sumes control of the government. However, in some societies, normally close to
the totalitarian pole, with no legacies of liberal or democratic politics, top
nomenklatura figures are able to put on nationalist garb and engage not in de­
mocracy building but ethnocracy building. In such contexts civil society is too
weak to generate a competitive political society and members of the nomen­
klatura are able to appropriate power and “legitimate” themselves via elections.
In relation to behavioral incentives, Communists (or ex-Communists) from
the former nomenklatura after defeat in free and contested elections will still oc­
cupy numerous important positions within the state apparatus, especially in state
enterprises. The members of the former nomenklatura through their networks
extending over management, administration, and even security services can as­
sure themselves a privileged position in the emerging capitalist economy and with
it substantial political influence. However, they normally act for their own indi­
70 Theoretical Overview

vidual self-interest. In most post-Communist countries the former nomenklatura

do not attempt to overthrow or directly challenge the new regime but to profit by
it. In some cases, particularly in the former Soviet Union, this leads to a confusion
between the public and the private and with it considerable room for corruption.
The more the members of the former nomenklatura act as individuals or demo-
cratic state managers, the better their chances of survival as officials. This is par­
ticularly so for managers of state production, trading, and banking enterprises,
who can use their organizational resources profitably to restructure new forms
of recombined public-private property.3 The incentive system for the former
nomenklatura thus has strong individualist or network components, which in­
volve working for advantages by manipulating the new political context more
than opposing it per se. The incentive system for the military is fundamentally
different. With few exceptions, incentives to the military are collective and derive
from the struggles to retain group prerogatives to avoid collective negative ac­
tions, such as trials. Therefore, unlike the nomenklatura out of office, for the mil­
itary out of office there may be significant incentives for acting together in open
contestation against the new democratic government.

Sultanistic Leadership
Last, we should briefly consider what the institutional composition of sul­
tanistic rule implies for democratic transitions and consolidation. A sultanistic
regime is one in which the ruler personalizes the government and the regime and,
in an uninstitutionalized but erratically pervasive way, penetrates the state, polit­
ical society, and civil society. Fused are not only the private and the public, but
also the civilian and the military. Theoretically, it is hard to classify sultanship as
either a military- or a civilian-led regime. Sultanistic regimes present an oppor­
tunity for democratic transition because, should the ruler (and his or her family)
be overthrown or assassinated, the sultanistic regime collapses. However, the very
nature of a sultanistic regime means that there is very little space for the organi­
zation of a democratic opposition. Therefore, short of death by natural causes,
sultanistic dictators are characteristically overthrown by quick, massive move­
ments of civil society, by assassination, or by armed revolt (see table 4.2). This
manner of regime termination often leads to the dynamics of a provisional gov­
ernment which, unless there is a decision to hold rapid elections, normally pre­
sents dangers for democratic consolidation.4 Also, the very personalization of
power around the dictator may allow close associates of the regime to assume
power. Or, even when the group or armed movement leading the revolt eliminates

3. Pioneering work on new network formation and the associated phenomenon o f “ recombinant prop­
erty” that is not really private and no longer public is being done by David Stark, “ Recombinant Property
in East European Capitalism,” Working Paper, Collegium Budapest, 1994.
4. We will discuss interim governments in our analysis o f the next variable.
 Actors and Contexts 71

those most associated with the sultanistic regime, they may appoint themselves as
the “ sovereign” representatives of the people and rule in the name of democracy
without passing through the free contestation and free election phases that are
necessary for full democratic transition and consolidation.

T r a n s it io n In it ia t io n : W ho S t a r t s a n d W ho C o n t r o l s ?

Transitions initiated by an uprising of civil society, by the sudden collapse of

the nondemocratic regime, by an armed revolution, or by a nonhierarchically led
military coup all tend toward situations in which the instruments of rule will be
assumed by an interim or provisional government.5 Transitions initiated by hier­
archical state-led or regime-led forces do not.
Interim governments are highly fluid situations and can lead to diametrically
opposite outcomes depending on which groups are most powerful, and especially
on whether elections or sweeping decree reforms are considered to be the first pri­
ority. If the interim government quickly sets a date for elections and rules as a rel­
atively neutral caretaker for these elections, this can be a very rapid and effica­
cious route toward a democratic transition. However, if the interim government
claims that its actions in overthrowing the government give it a legitimate man­
date to make fundamental changes that it defines as preconditions to democratic
elections, the interim government can set into motion a dangerous dynamic in
which the democratic transition is put at peril, even including the postponement
of elections sine die.
Elections are crucial because without them there is no easy way to evaluate
whether the interim government is or is not actually representing the majority.
Without elections, actors who did not play a central role in eliminating the old
regime will find it very difficult to emerge and assert that they have a democratic
mandate. And without elections the full array of institutions that constitute a new
democratic political society— such as legislatures, constituent assemblies, and
competitive political parties— simply cannot develop sufficient autonomy, legal­
ity, and legitimacy.
Elections are most likely to be held quickly in cases of collapse where demo­
cratic party leaders (as in Greece in 1974) almost immediately emerge as the core
of the interim government or where leaders of civil society who are committed to
creating a political democracy as the first order of business (as in Czechoslovakia
in 1989) are the core of the interim government. Frequently, however, especially in
cases where armed force has brought them into power, interim governments de­
velop a dynamic that moves them away from fully free contestation. Claiming

5. For a more detailed discussion o f interim governments, see Yossi Shain and Juan J. Linz, eds., Between
States: Interim Governments and Democratic Transitions (New York: Cambridge University Press, 1995).
 Theoretical Overview

revolutionary legitimacy, the provisional government may substitute occasional

plebiscites or referenda for multiparty elections. A provisional government that
begins with a nonhierarchical coup may open up an explosive situation because
it may involve part of the state apparatus attacking another part of the state ap­
paratus, in which outcomes can vary from massive state repression to revolution.
The least likely outcome in such a conflict is procedural democracy.
What can we say about state-led or regime-initiated and regime-controlled
transitions? For one thing the potential for the emergence of an interim govern­
ment is virtually absent when the regime controls the transfer of government
until elections decide who should govern. This fundamental point made, we need
to be aware that regime-controlled transfers can be placed along a continuum
ranging from democratically disloyal to loyal. A democratically disloyal transfer
is one in which, for whatever reasons, the outgoing regime attempts to put strong
constraints on the incoming, democratically elected government by placing sup­
porters of the nondemocratic regime in key state positions and by successfully in­
sisting on the retention of many nondemocratic features in the new political sys­
tem. A disloyal transfer is most likely to happen when the leaders of the outgoing
nondemocratic regime are reluctant to transfer power to democratic institutions
and the correlation of forces between the nondemocratic regime and the demo­
cratic opposition is one where the nondemocratic leaders retain substantial coer­
cive and political resources. For reasons we have already discussed, this is most
likely to happen if the prior nondemocratic government was a hierarchically con­
trolled military regime with strong allies in civil and political society, as we shall
see in the case of Chile.

In t e r n a t io n a l In f l u e n c e

The most influential and widely read publication on democratic transitions is

the four-volume work edited by Guillermo O’Donnell, Philippe C. Schmitter, and
Laurence Whitehead, Transitions from Authoritarian Rule. The cases in this study
all concerned Southern Europe and Latin America and, with the exception of
Italy, the decade of the mid-1970s to the mid-1980s. Generalizing from the expe­
riences within these spatial and temporal confines, O’Donnell and Schmitter in
the concluding volume argue that “domestic factors play a predominant role in
the transition. More precisely, we assert that there is no transition whose begin­
ning is not the consequence— direct or indirect— of important divisions within
the authoritarian regime itself.”6 Laurence Whitehead, in his valuable chapter on

6. Guillermo O’Donnell and Philippe C. Schmitter, Tentative Conclusions about Uncertain Democracies
(Baltimore: Johns Hopkins University Press, 1986), 19.
 Actors and Contexts 73

international influence, offers a more qualified generalization: “ In all the peace­

time cases considered here internal forces were of primary importance in deter­
mining the course and outcome of the transition attempt, and international fac­
tors played only a secondary role.”7
However, if one considers the entire world and all major actual (or potential)
cases of democratization in modern times, the analysis of international influences
can be pushed much further and a series of nuanced hypotheses can be advanced.
To do so, we distinguish between the foreign policy, Zeitgeist, and diffusion effects.

Foreign Policies
Conceptually, foreign policies can have an influence on domestic contexts in
very different ways. To begin with, there are in fact three categories of situations
in which the use of force in foreign policy actually determines outcomes that re­
late to democracy. First, a nondemocratic country can use force to overthrow a
less militarily powerful democracy and either annex or occupy the country or in­
stall a nondemocratic puppet regime (e.g., Germany in Czechoslovakia in 1938).
Second, a nondemocratic regional hegemon (which can be a single country or a
community of countries acting collectively) can in its “outer empire” use military
force to reverse a successful democratizing revolutionary effort to overthrow a
nondemocratic regime (e.g., Hungary in 1956) or to reverse a liberalizing process
(e.g., Czechoslovakia in 1968). Third, a democratic country that is a victor in a war
against a nondemocratic regime can occupy the defeated country and initiate a
democratic transition by installation (e.g., Germany and Japan in 1945). However,
although foreign policies can have determinative force in the democratic transi­
tion phase, democratic consolidation in an independent country is ultimately de­
termined by domestic forces.
Another influence of foreign policy on democratic transition and consolida­
tion concerns what we might call gate opening to democratic efforts. Formal or in­
formal empires, largely responding to their own internal and geopolitical needs,
may open a previously closed gate to democratization efforts in subordinate re­
gimes. Whether there will be a democratic transition or not and whether this will
lead to democratic consolidation or not is predominantly domestically deter­
mined (e.g., most of the British Empire afterWorld War II, the Soviet bloc in East­
ern Europe in 1989).

7. Laurence Whitehead, “ International Aspects o f Democratization,” in O’Donnell, Schmitter, and

Whitehead, Transitions from Authoritarian Rule: Comparative Perspectives, 4. In the body o f the article
Whitehead gives detailed information about how the European Community played a strongly supportive
role in democratic consolidation in southern Europe. In later works, Whitehead, O Donnell, and Schmit­
ter correctly acknowledged that international influence played a central role in Eastern Europe. Also see the
two-volume work edited by Abraham F. Lowenthal, Exporting Democracy: The United States and Latin
America (Baltimore: Johns Hopkins University Press, 1991); and Geoffrey Pridham, ed., Securing Democ­
racy: The International Context o f Regime Transition in Southern Europe (London: Routledge, 1990).
74 Theoretical Overview

Subversion is another kind of policy effect. Regional hegemons (democratic or

nondemocratic) can play an important contributing, though seldom determina­
tive, role in helping to subvert a nondemocratic regime (e.g., U.S. foreign policies
toward the Philippines in 1987) or in helping to subvert democracy that is oppos­
ing the hegemons policy preferences (e.g., U.S foreign policy toward Chile in
1973)- A democratic hegemon may also use its geopolitical and economic power
to thwart nondemocratic forces trying to impede a democratic transition process
(e.g., President Carter’s role in reversing electoral fraud in the Dominican Re­
public in 1978).
Finally, a regional hegemon may, by a consistent policy package of meaningful
incentives and disincentives, play a major supportive (but not determinative) role
in helping a fledgling democracy in the region complete a democratic transition
and consolidate democracy (e.g., the collective foreign policy of the European
Economic Community [EEC] and especially of West Germany toward Portugal
in 1974).

Zeitgeist
The concept of Zeitgeist is taken from the German tradition of intellectual his­
tory and refers to the “spirit of the times.” We do not believe in any variant of the
“end of history” thesis— the thesis, namely, that one ideology, such as the demo­
cratic ideology, can or will stop human efforts to respond to problems by creat­
ing alternative political visions and ideologies.8 But we do maintain that, when a
country is part of an international ideological community where democracy is
only one of many strongly contested ideologies, the chances of transiting to and
consolidating democracy are substantially less than if the spirit of the times is one
where democratic ideologies have no powerful contenders. The effect of a demo­
cratically hostile or a democratically supportive Zeitgeist can readily be seen when
we contrast interwar Europe with the Europe of the mid-1970s and the 1980s. In
interwar Europe, in the aftermath of the break-up of the Austro-Hungarian Em­
pire, boundary changes emerging out of the Treaty of Versailles, and various po­
litical experiments, eleven states with little or no prior experience of an indepen­
dent democratic regime made some effort to establish democracies.9 However,
the spirit of the times was one in which the democratic ideal competed with four

8. See, for example, Francis Fukuyama, “ The End o f History,” National Interests (Summer 1989): 3-18.
The return to power in Lithuania, Poland, and Hungary o f reform communists as social democrats is but
one example of how history can evolve in new and unexpected ways. Another example is the resurgence, in
the name o f “democratic majoritarianism,” o f ethnic nationalist dictatorships in parts o f the former Soviet
Union and Yugoslavia.
9. These states were Spain, Italy, Poland, Hungary, Czechoslovakia, Estonia, Latvia, Lithuania, Bulgaria,
Yugoslavia, and Romania. For a discussion o f their demise, see Juan J. Linz, “ La crisis de las democracias,”
in Mercedes Cabrera, Santos Julia, and Pablo Martin Acena, eds., Europa en crisis, 1919-1939 (Madrid: Edi­
torial Pablo Iglesias, 1992), 231-80.
 Actors and Contexts 75

other contesting ideologies in Europe, none of them democratic. Communism in

the Soviet Union was a novel experiment that many felt offered great promise.
Fascism in Italy was seen by many others as a powerful contestant to both com­
munism and democracy. Catholicism, after the papal encyclical, Rerum Novarum,
was the basis of novel forms of corporatist and integralist movements. Finally, in
the midst of this intense ideological struggle, many conservatives still remem­
bered positively the political formula of a predemocratic, authoritarian constitu­
tional monarchy, of which Imperial Germany was the esteemed exemplar. All of
Europe was influenced in some degree by these nondemocratic ideas. Latin Amer­
ica too was strongly influenced by these European intellectual and ideological
currents, as the experience of the Estado Novo under Vargas in Brazil and of Per-
onism in Argentina shows.
Though democracy is never ‘ overdetermined,” even in the context of the most
supportive Zeitgeist, by the late 1970s the Zeitgeist in southern Europe— indeed in
most of the world (with the important exception of a reinvigorated fundamental­
ism in the Islamic cultural community)— was such that there were no major ide­
ological contestants to democracy as a political system. To be sure, Communism
was entrenched in the Soviet Union and by extension in the subordinate regimes
of Eastern Europe, but the pronouncement by an eminent Polish philosopher that
the 1968 Soviet invasion of Czechoslovakia represented the “clinical death” of
Marxist revisionism in Central and Eastern Europe proved prophetic.10 By 1977,
the issue of human rights had acquired such pan-European support that most of
the East European regimes became signatories to the Helsinki Accords.11 Fascism
and Nazism were thoroughly discredited after World War II, and no longer repre­
sented a pole of attraction. After Vatican II (1961-63) Catholicism developed an
ideological and institutional position more amenable to democracy (if not to cap­
italism) than ever before.12 In the modern era most of the secure and successful
monarchs are now constitutional heads of state in parliamentary democracies. The
Egyptian and Peruvian military option so intriguing in the 1960s had few adher­
ents in the world by the mid-1970s. On the other hand, the Latin American left’s
experience with a new type of modern military-led bureaucratic-authoritarian
regime had contributed to a deep revalorization of democracy, not merely as a tac­
tical instrument but as a value in itself.13 The hopes that some democrats had in
Yugoslav worker self-management as a school for democracy have been thor­

10. Leszek Kolakowski, Main Currents o f Marxism (Oxford: Oxford University Press, 1978), 3: 4 ^5-
11. For the effects on the domestic politics o f East European countries and the Soviet Union o f having
signed the Helsinki Accords, see Samuel P. Huntington, The Third Wave: Democratization in the Twentieth
Century (Norman: Oklahoma University Press, 1991), esp. 85-100.
12. For Vatican II and how it enhanced the status o f democracy in Roman Catholic theology, see George
Weigel, The Final Revolution: The Resistance Church and the Collapse of Communism (New York: Oxford
University Press, 1992), esp. 67-74.
13. The revalorization o f democracy by the left produced a rich new genre o f writings. For one such ex­
ample see Francisco Weffort, “ Why Democracy?” in Alfred Stepan, ed., Democratizing Brazil: Problems of
Transition and Consolidation (New York: Oxford University Press, 1989), 327-50.
 Theoretical Overview

oughly disappointed. In Africa, “one-party” states by the early 1990s had lost al­
most all their original credibility as “mobilizing regimes” and were increasingly
disdained as “rent-seeking” formulas exploited by nondemocratic elites.

Diffusion
Zeitgeist in the world of politics refers to historical eras. But the diffusion effect
in an international political community, especially in a community tightly coupled
by culture, coercive systems, and/or communication, can refer to weeks or even
days. Law-like statements about human creations such as democracies are inher­
ently different from law-like statements in the physical sciences because no two
moments in history can be exactly alike. Human beings reflect upon previous
events and, where the events seem directly relevant to them, often consciously or
unconsciously attempt to adjust their behavior so as to achieve or avoid a com­
parable outcome. Political learning is possible. For example, after the Portuguese
revolution had exploded, a Spanish conservative leader, Manuel Fraga, expressed
some interest in playing a role in leading democratic change because he “did not
want to become the Caetano of Spain.” 14 Likewise Prince Juan Carlos in Spain
was undoubtedly influenced by the Greek case, where his brother-in-law, King
Constantine, lost his throne due to his ambivalence about democracy.
More generally, we posit that the more tightly coupled a group of countries
are, the more a successful transition in any country in the group will tend to
transform the range of perceived political alternatives for the rest of the group.
Indeed, as we shall see when we examine Central and Eastern Europe in 1989, in­
ternational diffusion effects can change elite political expectations, crowd be­
havior, and relations of power within the regime almost overnight. For practi­
tioners and theorists alike, diffusion effects have obviously gained in salience in
the modern world owing to the revolution in communications. Today, the dra­
matic collapse of a nondemocratic regime is immediately experienced by virtu­
ally the entire population of the neighboring countries through radio and televi­
sion. This experience in turn instantly becomes a powerful new component of
domestic politics.15

T h e P o l it ic a l E c o n o m y of L e g it im a c y a n d of C o e r c io n

What is the relationship between citizens’ perception of the socioeconomic

efficacy of a regime and their perception of the legitimacy of the regime itself?

14. Fraga was referring to the overthrow o f the post-Salazar leader o f Portugal, Marcello Caetano, who
failed to initiate a transition. The diffusion effect here is that Spanish conservatives rapidly began to recal­
culate the costs and benefits o f initiating a democratic transition.
15. All countries discussed in this volume experienced some diffusion effects, but none more dramati­
cally than the countries o f Central and Eastern Europe.
 Actors and Contexts 77

How does the economy affect the prospects of a transition away from a nondem-
ocratic regime? If a transition has begun, how does the economy affect the
chances of democratic consolidation? Are democratic and nondemocratic regimes
equally helped by sustained growth? Are democratic and nondemocratic re­
gimes equally hurt by economic decline?
We accept the well-documented correlation that there are few democracies at
very low levels of socioeconomic development and that most polities at a high
level of socioeconomic development are democracies.16 Most of the major mod­
ern transition attempts thus take place in countries at medium levels of develop­
ment. However, this relationship between development and the probability of de­
mocracy does not tell us much about when, howyand if a. transition will take place
and be successfully completed. Indeed, within this critical context of intermedi­
ate levels of development we contend that it is often difficult or impossible to
make systematic statements about the effect of economics on democratization
processes.17 However, if one uses an analytical framework that combines politics
and economics and focuses on legitimacy, one can make much more meaningful
statements. Certainly for transition theory, economic trends in themselves are less
important than is the perception of alternatives, system blame, and the legitimacy
beliefs of significant segments of the population or major institutional actors.
Why?
For theoreticians and practitioners who posit a tightly coupled relationship
between the economy and regime stability, robust economic conditions would
appear supportive of any type of regime. We would argue, however, that the
proposition is theoretically and empirically indefensible. We see good theoretical
reasons why sustained economic growth could erode a nondemocratic regime.
We see no theoretical reason why sustained economic growth would erode a dem­
ocratic regime. Regime type can make a great difference. From the perspective of
political economy, we absolutely cannot formulate any valid propositions that
take the form, “under conditions of great economic prosperity there will be no in­
centives for a transition from a nondemocratic to democratic regime.” This is so
precisely because many nondemocratic regimes, especially those of the statistical

16. The classic initial formulation o f this argument was Seymour Martin Upset, “Some Social Requi­
sites o f Democracy: Economic Development and Political Legitimacy,” American Political Science Review
(March 1959): 69-105. Larry Diamond reviewed three decades of literature relevant to the development/de­
mocracy debate and concluded that the evidence broadly supports the Lipset theory. See Diamond, Eco­
nomic Development and Democracy Reconsidered,” in Gary Marks and Larry Diamond, eds., Reexamin­
ing Democracy (Newbury Park: Sage, 1992), 93-139.
17. The specific relationship between economic growth or economic crisis and the initiation o f a tran­
sition out o f a nondemocratic regime has been the object o f considerable debate. José Maria Maravall, in
an outstanding and well-researched work, has analyzed this problem in great detail, with particular refer­
ence to southern and Eastern Europe. We find that his analysis converges with our brief analysis, which we
had written independently. We are happy to refer the reader to his book for the relevant evidence. See José
Maria Maravall, Los resultados de la democracia: Un estudio del sur y el este de Europa (Madrid: Alianza
Editorial, 1995).
78 Theoretical Overview

mode, authoritarian regimes, are originally defended by the state elite and their
core socioeconomic allies as necessary given the exceptional difficulties (often
economic) the polity faces. Thus, prolonged economic prosperity, especially in an
authoritarian regime, may erode the basis of the regimes justification based on
exceptional circumstances. Prolonged economic success can contribute to the
perception that the exceptional coercive measures of the nondemocratic regime
are no longer necessary and may possibly erode the soundness of the new eco­
nomic prosperity.
Prolonged economic growth may also contribute to social changes that raise
the cost of repression and thus indirectly facilitate a transition to democracy. Pro­
longed economic expansion normally contributes to the growth of a middle class;
a more important and needed skilled labor force; an expansion of education;
greater contacts with other societies via television, radio, and travel; and a more
diverse range of possible protests. There is even strong evidence to indicate that,
within a territory, increases in regional wealth increase citizens’ expectations that
they should be well treated by the police.18
Empirically, there are a number of cases where sustained prosperity altered re­
lations of power in favor of democratic forces. In fact, three cases in our study,
Pinochets Chile, Brazil in the early 1970s, and Franco’s Spain in its last twenty
years (as well as South Korea), had some of the world’s highest rates of economic
growth. Spain’s growth contributed to the belief of some of the core constituents
of the authoritarian regime and among the industrial elite that they could man­
age equally well in the future in a more democratic environment. The times had
changed and so did the regime.19 In Brazil, the soft-line military wing announced
its liberalization program in September 1973, after five years of unprecedented
growth and before the oil crisis, soaring interest rates, and its attendant debt cri­
sis. In September 1973 the military felt that the economy was in excellent condi­
tion and no significant political threat existed. In the absence of the “exceptional
circumstances” that had legitimated their coup in their own eyes, they came to be­
lieve that continued authoritarian rule not only was not necessary but might con­
tribute to the autonomy of the security forces and the “Argentinization of
Brazil.”20 In Chile many of the key industrialists who had believed that Pinochet

18. For example, seven occupational groups in Franco’s Spain, ranging from manual laborers to those
in liberal professions, were asked if they expected “ equal,” “ better,” or “worse” treatment by the police than
other citizens. The data were broken down according to the level o f economic development o f the respon­
dents’ place o f residence. In 19 of 21 o f the possible comparisons, the greater the regional economic devel­
opment, the greater the expectation o f equal treatment by the police. See Juan J. Linz, “ Ecological Analysis
and Survey Research,” in Mattei Dogan and Stein Rokkan, eds., Quantitative Ecological Analysis in the So­
cial Sciences (Cambridge: M IT Press, 1969)* 91-131. esp. table 1, p. 113.
19. As Adolfo Suárez said before he became prime minister o f Spain, “Our people who at the beginning
of his (Franco’s) government had asked simply for bread, today ask for quality consumption, and in the
same fashion, whereas at the beginning they wanted order, today they ask for freedom— freedom o f polit­
ical association.” Speech in the Cortes on June 9,1976.
20. Stepan, Rethinking Military Politics, 32-33.
 Actors and Contexts 79

was indispensable in 1980, by 1988 had come to believe that the risk of fair elec­
tions to the economic model was less than the risk of supporting Pinochet in un­
fair elections.21 In all three cases, the political economy of prosperity contributed
to new perceptions about alternative futures and to lessening resistance to demo­
cratic alternatives.
In sharp contrast, when we consider democratizing regimes or consolidated
democracies, there are no theoretical reasons or empirical evidence to support an
argument that economic growth contributes to regime erosion. Of course, a “ rev­
olution of rising expectations” may create new demands on democratic govern­
ments, but it cannot attack their raison d'être. Indeed, if a regime is based on
the double legitimacy of democratic procedures and socioeconomic efficacy, the
chances of a fundamental regime alternative (given the absence of a “stateness”
problem) being raised by a significant group in society is empirically negligible.
Severe economic problems affect democratic and nondemocratic regimes, es­
pecially authoritarian ones, very differently. There are good theoretical reasons
why sharp economic decline (say five years of continuous negative growth) will
adversely affect stability in both democratic and nondemocratic regimes, but it
will affect the latter substantially more. Modern nondemocratic (especially au­
thoritarian) regimes are often heavily dependent on their performance claims but
are not bolstered by procedural claims deriving from their democratic status.
Theory leads us to posit therefore that a democratic regime has two valuable
sources of insulation from sustained economic downturn not available to a non­
democratic regime: its claim to legitimacy based on its origin and the fact that
elections are always on the horizon and hold the prospect of producing an alter­
native socioeconomic program and an alternative government without a regime
change. This means that most new democracies have about eight years of breath­
ing space— four years or so for the initial government and four years or so for an
alternative government.
This theory-based assumption gains strong empirical support from data com­
piled by Fernando Limongi and Adam Przeworski. In their study of South Amer­
ica between 1945 and 1988, they found that the probability that a nondemocratic
regime would survive three consecutive years of negative growth was 33 percent,
whereas the probability that a democratic regime would survive three years of neg­
ative growth was 73 percent. More dramatically, their data show that no nondem­
ocratic regime survived more than three years of consecutive negative growth,
whereas the probability that a democratic regime would survive four or five years
of consecutive negative growth was 57 percent and 50 percent respectively.22
Let us return to our argument concerning economics and the politics of alter­

21. See the interview with one o f the leaders o f a major business interest group in Chile, in Alfred
Stepan, “ The Last Days o f Pinochet?” New York Review o f Books (June 2,1988): 33.
22. Fernando Limongi and Adam Przeworski, “ Democracy and Development in South America,
1945-1988” (University o f Chicago, October 27,1993, unpublished manuscript).
8o Theoretical Overview

natives and system blame in nondemocracies and in democracies. If the political

situation is such that there is no strong perception of a possible alternative, a non-
democratic regime can often continue to rule by coercion. However, when the be­
lief grows that other alternatives are possible (as well as preferable), the political
economy of legitimacy and coercion changes sharply. If the coercive capacity of
the nondemocratic regime decreases (due say to internal dissent or the with­
drawal of vital external guarantees), then the political economy of prolonged
stagnation can contribute to the erosion of the regime. It is not changes in the
economy, but changes in politics, that trigger regime erosion— that is, the effects
of a poor economy often have to be mediated by political change.
The question of system blame is also crucial for the fate o f democracies. As we
have discussed elsewhere, the economic crisis of interwar Europe was as intense
in countries such as the Netherlands and Norway (which did not break down)
as in Germany and Austria (which did break down). Indeed, 30,000 Dutch work­
ers in 1936 went to work in Germany because the Dutch economy was in worse
condition than the German economy. What made the crisis of the economy a cri­
sis of the political system in Germany and Austria was that strong groups on the
right and the left had regime alternatives in mind and thus attacked the regime.
Politically motivated system blame, more than the economic crisis per se, caused
the German and Austrian breakdowns.23
The key question for the democracies is whether their citizens believe that, in
the circumstances, the democratic government is a doing a credible job in trying
to overcome economic problems. It is important to stress that the political econ­
omy of legitimacy will produce severe and perhaps insoluble challenges to dem­
ocratic consolidation in those cases where the democratic system itself is judged
to be incapable of producing a program to overcome the economic crisis.
To summarize, what can and cannot we say about transition theory and the
political economy of legitimacy? Theory and the Limongi-Przeworski data indi­
cate that consecutive years of negative growth lessen the chance of either a non­
democratic or a democratic regimes surviving. Thus, a country that is experi­
encing positive growth, other things being equal, has a better chance to
consolidate democracy than a country that is experiencing negative growth. This
said, the theory and the data also indicate that a democratic regime has more in­
sulation from economic difficulty than does a nondemocratic regime. The ques­

23. For a more detailed development o f this argument with supporting data, see Juan J. Linz and Alfred
Stepan, “ Political Crafting o f Democratic Consolidation or Destruction: European and South American
Comparisons,” in Robert A. Pastor, ed., Democracy in the Americas: Stopping the Pendulum (New York:
Holmes and Meyer, 1989), 41-61. We are indebted to Ekkart Zimmerman for his pioneering studies o f in­
terwar Europe. See Zimmerman, “ Government Stability in Six European Countries during the World Eco­
nomic Crisis o f the 1930s: Some Preliminary Considerations,” European Journal o f Political Research 15, no.
1 (1987): 23-52 and Zimmerman, “ Economic and Political Reactions to the World Economic Crises o f the
1930s: Six European Countries,” paper presented for the Mid-West Political Science Association Conven­
tion, Chicago, April 10-12,1986.
 Actors and Contexts 81

tion of whether an aspiring democracy can withstand economic difficulties, as the

German-Dutch comparison showed, depends to a great extent on the degree of
noneconomic system blame and mass-elite perceptions about the desirability of
other political alternatives. The question is thus one of relationships. It is theo­
retically possible, and indeed has occurred, that a newly democratizing regime
suffers a decline in citizen perceptions of democracy’s socioeconomic efficacy at
the same time that their belief that “democracy is the best possible political system
for a country like ours” increases.24
In those cases, however, where the citizens come to believe that the democratic
system itself is compounding the economic problem or is incapable of defining
and implementing a credible strategy of economic reform, system blame will
greatly aggravate the political effect of economic hard times. More importantly,
economic crises will tend to lead to democratic breakdown in those cases where
powerful groups outside or— more fatally— inside the government increasingly
argue that nondemocratic alternatives of rule are the only solution to the eco­
nomic crisis.
In a situation where the crisis is permanent, after at least one democratic al­
ternation of government, and where a reasonable argument can be made that the
democratic political actors are incapable or unwilling to search for solutions and
even compound the problems by such actions as infighting and corruption, key
actors will search for alternatives. But alternatives might not be available. Key ac­
tors’ previous experience with alternatives might have been equally or more un­
attractive. In such circumstances, many of these actors might resign themselves to
a poorly performing democracy. Such resignation may not prevent crises, up­
heavals, and attempted local coups but is not conducive to regime change. But it
certainly makes consolidation difficult and can even deconsolidate a democracy.

C o n s t it u t io n -M a k in g E n v ir o n m e n t s

A neglected aspect of democratic transition and consolidation concerns the

comparative analysis of the contexts in which constitutional formulas are
adopted or retained. Without attempting to review all possible variations, let us
simply mention six very different possible constitution-making contexts and/or
formulas and indicate what problems they present for democratic transition and
democratic consolidation. We move from those contexts and formulas that pre­

24. In Linz and Stepan, “ Political Crafting o f Democratic Consolidation or Destruction, 44, we note
that, during a period (1978-1981) o f rising unemployment, inflation, recession, and terrorism the Spanish
citizens belief in the efficacy o f democracy declined by 25 percentage points in national polls while the be­
lief that democracy was the best political system for a country like Spain increased by 5 percentage points
in the same period. The key implication is that the citizenry did not believe, despite the economic prob­
lems, that any alternative political system was preferable.
82 Theoretical Overview

sent the most confining conditions for democratic consolidation in an existing

state to those that present the least.25
1. The retention of a constitution created by an nondemocratic regime with
reserve domains and difficult amendment procedures. These confining condi­
tions may be the price the outgoing nondemocratic regime is able to extract for
yielding formal control of the state apparatus. However, if this constitution dejure
enshrines nondemocratic “ reserve domains” insisted upon by the outgoing non­
democratic power-holders, then the transition by our definition cannot be com­
pleted until these powers are removed. If the constitution has very difficult
amendment procedures this will further complicate the process of democratic
transition and consolidation. In this book Chile is the clearest case.
2. The retention of a “paper” constitution which has unexpected destabilizing
and paralyzing consequences when used under more electorally competitive con­
ditions. Some nondemocratic constitutions may enshrine a very elaborate set of
decision-rules, procedures, and rights that had no effect on the operation of the
nondemocratic regime because the constitution was a fiction. However, in more
electorally competitive circumstances, this constitution can take on a life of its own
that may make it almost impossible to arrive at democratically binding decisions.
In such cases, the constitution can help destroy the state and should be changed ex­
tremely quickly before its perverse consequences have this paralyzing effect. The
most important instances of this type of constitution are found in the Soviet-type,
federal constitutions in the former USSR, Yugoslavia, and Czechoslovakia.
3. The creation by a provisional government of a constitution with some de
jure nondemocratic powers. Even when the old nondemocratic regime is de­
stroyed and many new policies are passed, a democratic transition itself cannot
be completed unless the nondemocratic components of the constitution crafted
by the provisional government are eliminated, as we shall see in the case of Por­
tugal. Even when these nondemocratic clauses are eliminated, the origin of the
constitution in a provisional government may hurt democratic consolidation be­
cause of its inappropriateness or weak societal acceptance.
4. The use of constitution created under highly constraining circumstances
reflecting the de facto power of nondemocratic institutions and forces. Such a
constitution may be formally democratic and thus consistent with a transition

25. Some indispensable sources on constitutions and democracy are Jon Elster and Rune Slagstad, eds.,
Constitutionalism and Democracy (Cambridge: Cambridge University Press, 1988); Douglas Greenberg,
Stanley N. Katz, Melanie Beth Oliveira, and Steven C. Wheatly, eds., Constitutionalism and Democracy:
Transitions in the Contemporary World (New York: Oxford University Press, 1993); Bruce Ackerman, The
Future of Liberal Revolution (New Haven: Yale University Press, 1992); A. E. Dick Howard, ed., Constitution
Making in Eastern Europe (Washington: Woodrow Wilson Center Press, 1993); and the East European Con­
stitutional Review, published quarterly since 1992 by the Center for the Study o f Constitutionalism in East­
ern Europe at the University o f Chicago Law School in partnership with the Central European University.
 Actors and Contexts 83

being completed, but democratic consolidation may be hampered because a con­

strained constituent assembly, while believing that other institutional arrange­
ments are more appropriate for the creation and consolidation of democratic
politics, may be de facto prevented from selecting them. To some extent Brazil is
such a example.
5. The restoration of a previous democratic constitution. This formula pre­
cludes a potentially divisive debate about constitutional alternatives and is often
selected by redemocratizing polities for reasons of speed, conflict avoidance, and
the desire to call upon some legacies of historic legitimacy. It should be pointed
out, however, that simple restoration presents two potential problems for demo­
cratic consolidation. First, when the polity has undergone great changes during
the authoritarian interlude, it is possible that a new constitutional arrangement
would in fact be more appropriate for democratic consolidation. Second, restora­
tion also assumes that the political procedures and institutions of the old consti­
tution have played no role whatsoever in the democratic breakdown. When the
old democratic arrangements have in fact contributed to democratic breakdown,
restoration precludes an historic opportunity to construct new and improved
arrangements with different procedures and symbols. Uruguay and Argentina are
cases worth analyzing from this perspective.
6. Free and consensual constitution-making. This occurs when democrati­
cally elected representatives come together to deliberate freely and to forge the
new constitutional arrangements they consider most appropriate for the consol­
idation of democracy in their polity. The constituent assembly ideally should
avoid a partisan constitution approved only by a “temporary majority” that leads
a large minority to put constitutional revisions on the agenda, thereby making
consolidation of democratic institutions more difficult. The optimal formula is
one in which decisions about issues of potentially great divisiveness and intensity
are arrived at in a consensual rather than a majoritarian manner and in which the
work of the constituent assembly gains further legitimacy by being approved in a
popular referendum that sets the democratic context in which further changes,
such as devolution (if these are to be considered), take place.26 In this book only
Spain fits this pattern.
In the rest of this book we examine how the interplay of our arenas, such as po­
litical society, rule of law, and economic society, and our variables, such as regime
type, stateness, and those discussed in this chapter, affected the processes of tran­
sition to democracy and the consolidation of democracy in three different so­
ciopolitical (and geographic) regions of the world— southern Europe, the South­
ern Core of Latin America, and post-Communist Europe.

26. For an argument in favor o f consensual constitutions produced and ratified by nationwide debates,
see Ackerman, The Future o f Liberal Revolution, 46-68.