Alfred Stepan, Juan J. Linz, Yogendra Yadav - Crafting State-Nations. India and Other Multinational Democracies PDF
Alfred Stepan, Juan J. Linz, Yogendra Yadav - Crafting State-Nations. India and Other Multinational Democracies PDF
Alfred Stepan, Juan J. Linz, Yogendra Yadav - Crafting State-Nations. India and Other Multinational Democracies PDF
Crafting State-Nations
India and Other Multinational Democracies
alfred stepan
juan j. linz
yogendra yadav
Stepan, Alfred C.
Crafting State-Nations : India and other multinational democracies /
Alfred Stepan, Juan J. Linz, and Yogendra Yadav.
p. cm.
Includes bibliographical references and index.
ISBN-13: 978-0-8018-9723-8 (hardcover : alk. paper)
ISBN-10: 0-8018-9723-8 (hardcover : alk. paper)
ISBN-13: 978-0-8018-9724-5 (pbk. : alk. paper)
ISBN-10: 0-8018-9724-6 (pbk. : alk. paper)
1. Multinational states—Case studies. 2. Democracy—Case studies.
I. Linz, Juan J. (Juan José), 1926– II. Yadav, Yogendra. III. Title.
JC311.S827 2010
321%.8—dc22 2010006887
A catalog record for this book is available from the British Library.
Special discounts are available for bulk purchases of this book. For more
information, please contact Special Sales at 410-516-6936 or
specialsales@press.jhu.edu.
Bibliography 277
Index 297
Tables and Figures
tables
figures
The territory of the world today is divided into 195 states; 192 of them are members
of the United Nations, and their boundaries are internationally recognized by
other states.1 Such internationally recognized states are presumed to have the
right to exercise authority over the population within their borders, whether the
people are citizens, subjects, or even foreigners. In many cases, this authority has
little to do with the population having a ‘‘we-feeling’’ as members of a community
or as members of a nation, because the states were not created by a coherent
nation, instead arising as the result of rulers successfully imposing themselves,
often by wars or international settlements following wars. No new independent
state was created by a national movement without some other existing states
supporting it and without very important states, normally the international system
of states, recognizing it. What we see as ‘‘nation-states’’ all have had a major
component of being ‘‘constructed’’ by existing powers. A nation-state without a
prior state helping construct it is inconceivable. The major functional alternative
to a nation-state is what we will call a ‘‘state-nation.’’ But note: both ‘‘nation-states’’
and ‘‘state-nations’’ are states, and must be states if they are to work.
This leads us to the major theme of this book, which is ‘‘state-nations.’’ This
may seem an awkward term. We have considered many alternatives, but we keep
returning to ‘‘state-nations,’’ because both the state and the nation are indispens-
ible elements for modern democracies but stand in an opposite relationship
to each other in our state-nation model than they do in the standard nation-
state model.
Democracy entails the democratic management of a specific territorial state
and its citizens. For too long, the normatively privileged model for a modern state
1. The Vatican is recognized as a state by many states but is a not a member of the United
Nations. Kosovo is recognized by the United States, but it is still not a member of the United
Nations. Taiwan is not a member state of the United Nations but even the divided states North
Korea and South Korea are.
xii preface
has been the nation-state. However, in some countries, more than one group
thinks of itself as a nation and has leaders who strive for independence. In this
book, we call such states ‘‘robustly politically multinational.’’ We are convinced
that in some circumstances, especially if a polity is ‘‘robustly multinational,’’ a
politics of nation-state-building is in conflict with a politics of inclusionary de-
mocracy and societal peace. In our judgment, therefore, the complexities, con-
flicts, and identities of citizens require the theoretical, normative, and political
imagining of alternatives to the nation-state model.
In this book, our major alternative is what we call the state-nation model. In
chapter 1, we present the core assumptions of the standard nation-state model in
Weberian ideal-type terms and then create a normatively and institutionally co-
herent alternative ideal type, the state-nation model, and show how it stands in
sharp contrast to the nation-state’s core assumptions.2 Since we are interested in
realizable and observable political alternatives, we then propose a ‘‘nested set’’ of
six policies that we believe are supportive of the crafting of state-nations.
We believe our model can, and should be, subject to empirical testing. We
thus stipulate that if a polity is close to a state-nation ideal type—even if the state
recognizes and supports numerous different languages, cultures, and indeed,
nations within the polity—its citizens should have four empirically documentable
characteristics. These four characteristics are: (1) a high degree of positive identi-
fication with the state; (2) multiple but complementary political identities; (3) a
high level of trust in the state’s institutions; (4) a high degree of positive support for
democracy among all the extremely diverse groups of citizens in the country.3
A central claim of nation-state theorists is that only the nation-state can culti-
vate the trust and identification with the state that a functioning democracy
requires. In an opening test of this hypothesis, we explore, using data from the
World Values Survey, the degree of trust in six key political institutions found
in the eleven longstanding federal democracies in the world. We divided these
eleven longstanding federations into those closest to the state-nation pole (Switzer-
land, Canada, Belgium, Spain, and India) and those states closest to the nation-
state pole (Germany, Austria, the United States, Australia, Argentina, and Brazil).
We find that states closer to the state-nation pole actually score higher on trust than
states closer to the nation-state pole. We also show that in Spain, even in regions
like Catalonia, where Catalan and Spanish are official languages and where there
2. For the reader who would like to see these two ideal types side by side in one table before
going further, please see table 1.1.
3. If a polity is close to what we call ‘‘pure multinational’’ because it has a series of virtual nation-
states within its borders, these attitudes will not be found. See chapter 1, especially figure 1.1.
preface xiii
4. CSDS uses booster samples in its nationwide surveys to ensure sufficient respondents from
smaller states (e.g., Mizoram) or social groups. They also do surveys devoted to a single state such as
Kashmir or Punjab.
5. The major exception to this is Kashmir, where respondents in the last decade have twice
indicated a preference for an independent state.
xiv preface
6. Sri Lanka also began its democratic experiment with greater per capita income and greater
literacy than India.
preface xv
One problem that the present work has left out as a central concern, except for
our small final chapter on the United States, is the complex relationships between
federalism and democracy, between federalism and fraternity, and especially be-
tween federalism and equality. This problem cannot be studied without com-
parativists addressing why the United States has some of the highest rates of
inequality in the democratic world. Inequality in the United States deserves
monographic study. Indeed, the great difficulty of passing fundamental welfare
legislation like health reform in the United States would seem to make this a
major subject for further work by us and other scholars and practitioners.
A final comment here on the long genesis of this book is in order. In 1995,
when Juan Linz and Alfred Stepan were completing their book Problems of Demo-
cratic Transition and Consolidation: Southern Europe, South America, and Post-
Communist Europe, they had already turned their attention to their next book, on
federalism, democracy, and nationalism.
Linz and Stepan had been drawn to this theme because they were aware of the
relative success of federalism in the case of Spain, on the one hand, and the
much less successful post-Soviet and Yugoslav experiences, on the other. Linz and
Stepan were also uneasy with the standard treatments of federalism in the litera-
ture and felt that they should aim at a new and more general theory.7 Linz and
Stepan agreed that one of the major theoretical and political problems of our
time was to conceptualize and realize political arrangements whereby deeply
diverse cultures, even different ‘‘nations,’’ can peacefully and democratically co-
exist within one state.
One of the most interesting cases of the successful solving of this problem
appeared to be India, and Stepan began to make almost yearly research visits to
that country starting in the late 1990s. During these visits, Stepan began to work
with Yadav, the founder and convener of the Lokniti Network. Yadav invited Linz
and Stepan to work with him in drafting questions for CSDS surveys.
As a result of the exciting findings that began to emerge from this process, Linz
and Stepan decided to abandon their idea of writing a theoretical and compara-
tive book that would have covered the entire world in a necessarily abstract way in
favor of a more focused, comparative analysis of key countries, situations, and
models, that could be empirically richer and based on their own original research.
7. Linz, as a social scientist and as a citizen of Spain, has been drawn to issues of nationalism
since the late 1960s. Indeed, the second volume of his seven volume collected works is devoted
entirely to nationalism and to federalism. See Juan J. Linz, ‘‘Obras Escogidas,’’ in Nación, Estado y
Lengua, ed. José Ramón Montero and Thomas Jeffrey Miley (Madrid: Centro De Estudios Políti-
cos y Constitutionales, 2008).
preface xvii
Linz and Stepan found the exchanges with Yadav, which opened up such rich
and often quite counterintuitive findings, so intellectually exciting that they ea-
gerly invited Yadav to join them in the writing of what has become this book. The
long and very collaborative process, which we hope has been fruitful for the book,
has certainly been fruitful for us.
This book has only been possible because of the support of dozens of col-
leagues, students, and organizations in many countries. Let us begin with our
students and colleagues. Stepan is fortunate to have had at Columbia a number of
brilliant Ph.D. candidates who shared his passion for examining many of the
puzzles in this book and who provided invaluable research assistance and in-
tellectual partnership. In particular he wants to recognize the contributions of
Neelanjan Sircar, Pavithra Suryanarayan, Israel Marquez, and Enrique Ochoa
Reza. Thomas Jeffrey Miley, the author of an important book on nationalism,
Nacionalismo y política lingüística: el caso de Cataluña, who is now on the faculty
at Cambridge University in the United Kingdom, helped all of us, particularly
Juan Linz, in developing the foundations of this book. While Miley was working
on his dissertation at Yale, he was involved in virtually all of the early discussions of
this book.
We thank the World Values Survey for opening their datasets to us. We are
grateful to the extraordinary scholars in the Lokniti Network whose support we
continuously drew upon, including Sanjay Kumar, Dhananjai Joshi, Sanjeer
Alam, and Himanshu Bhattacharya and Kanchan Malhotra of the CSDS Data
Unit for their help in accessing and analyzing the vast datasets at the CSDS. We
thank Rekha Chowdhary of the University of Jammu and G. K. Prasad of the
University of Madras for sharing their ideas and data to help us understand the
politics of their states. Suhas Palshikar, of the University of Pune, and Peter
deSouza, director of the Indian Institute of Advanced Study, were two of the
principal investigators of the State of Democracy in South Asia study, and we
thank them for sharing their insights and data with us. Yadav also thanks the
Wissenschaftskolleg zu Berlin, where he was a fellow and had an opportunity to
work on the final drafts of this book.
Many of the ideas in this book were presented to large groups of theorists and
practitioners, including at the United Nations Development Program where we
helped to conceptualize and write Human Development Report 2004: Cultural
Liberty in Today’s Diverse World, working closely with Amartya Sen.
The first joint article of Stepan, Linz, and Yadav was produced for a volume
edited by Ambassador Shankar Bajpai, titled Democracy and Diversity: India and
the American Experience. Bajpai brought together, over the course of a decade,
xviii preface
allowed us to explore in much greater depth than we otherwise could have, the
relationship of religions, democracies, and state-nations.
We dedicate this book to Rocío de Terán, Madhulika Banerjee, and Nancy
Leys Stepan, all of whom have their own projects but who always inspired and
helped us in our attempts to imagine better futures.
Crafting State-Nations
chapter one
One of the most urgent conceptual, normative, and political tasks of our day is to
think anew about how polities that aspire to be democracies can accommodate
great sociocultural, even multinational, diversity within one state. The need to
think anew arises from a mismatch between the political realities of the world we
live in and the old political wisdom that we have inherited. The old wisdom holds
that the territorial boundaries of a state must coincide with the perceived cultural
boundaries of a nation. This understanding requires that every state must contain
within itself one and not more than one culturally homogeneous nation, that
every state should be a nation, and that every nation should be a state. Given the
reality of sociocultural diversity in many of the polities of the world, this wide-
spread belief seems to us to be misguided, indeed dangerous, since, as we shall
argue, many states in the world today do not conform to this expectation.
While all independent democratic states have a degree of cultural diversity, for
comparative purposes we can say that, at any given time, states may be divided
analytically into three different categories:
1. States that have deep cultural diversity, some of which is territorially based
and politically articulated by significant groups that, in the name of na-
tionalism and self-determination, advance claims of independence;
2. States that are culturally quite diverse, but whose diversity is nowhere orga-
nized by territorially based politically significant groups mobilizing na-
tionalist claims for independence; and
3. States in which a community, culturally homogeneous enough to con-
2 crafting state-nations
sider itself a nation, dominates the state and no other significant group
articulates similar claims.
In this book, we will call countries, part of whose territory falls into the first
category, ‘‘robustly multinational’’ societies. Canada (owing to Quebec), Spain
(Basque Country and Catalonia), and Belgium (Flanders) are ‘‘robustly multi-
national.’’ India, due to the Kashmir Valley alone, merits classification in this
category. Furthermore, at various times the Mizo and Naga struggle for indepen-
dence in northeast India, the Khalistan movement in the Punjab, the once-
separatist Dravidian movement in southern India, and other secessionist move-
ments have also given a robust multinational dimension to Indian politics.
Switzerland and the United States are both sociologically diverse and multi-
cultural. However, since neither country has significant territorially based groups
mobilizing claims for independence, both countries clearly fall into the second
and not the first category.
Finally, countries such as Japan, Portugal, and most of the Scandinavian coun-
tries fall into the third category. It is not that these countries are devoid of ethnic
minorities and regional differences, but as of now these differences are politically
not very salient.
What political implications do these three very different situations have for
reconciling democracy with diversity? For us a major implication is that, if at the
time of the inauguration of competitive elections, a polity has only one significant
group that sees itself as a nation and there exists a relatively common sense of
history and religion and a shared language throughout the territory, nation-state
building and democracy building can be mutually reinforcing logics.
However, if competitive elections are inaugurated under conditions that are
already ‘‘politically robustly multinational,’’ nation-state building and democracy
building are conflicting logics. This is so because only one of a given polity’s
‘‘nations’’ would be privileged in the nation-state-building effort, and the others
would not be recognized or would even be marginalized. But before examining
alternatives to the nation-state, we first need to explain the normative and political
power of the nation-state.
The belief that every state should be a nation is perhaps the most widely accepted
normative vision of a modern democratic state, that is, the nation-state. After the
French Revolution, especially in the late nineteenth century, many policies were
comparative theory and political practice 3
1. For a classic book on these policies, see Eugen Weber, Peasants into Frenchman: The Modern-
ization of Rural France, 1870–1914 (Stanford: Stanford University Press, 1976). Most nineteenth-
century progressives and democrats, particularly those associated with the French Revolution, were
profoundly opposed to federalism. On the normative advocacy of a unified, homogeneous, nation-
state, see the entries on ‘‘Federalism,’’ ‘‘Federation,’’ ‘‘Nation,’’ and ‘‘Departement’’ in the extremely
illustrative but not widely known François Furet and Mora Ozouf, eds., A Critical Dictionary of the
French Revolution (Cambridge: Belknap Press, 1989), pp. 54–64, 65–73, and 742–753.
2. See Juan J. Linz, ‘‘State Building and Nation Building,’’ European Review 1 (1993), pp. 355–
369; and Juan J. Linz and Alfred Stepan, Problems of Democratic Transition and Consolidation:
Southern Europe, South America, and Post-Communist Europe (Baltimore: Johns Hopkins Univer-
sity Press, 1996), ch. 2. Also see Alain Fenet, ‘‘Difference Rights and Language in France,’’ in
Language, Nation, and State: Identity Politics in a Multilingual Age, ed. Tony Judt and Denis
Lacorne (New York: Palgrave Macmillan, 2004), pp. 19–62.
4 crafting state-nations
polity need to think about, craft, and normatively legitimate a type of polity with
characteristics of a ‘‘state-nation.’’
Linz and Stepan first introduced this concept in 1996, but only in a paragraph
(and one figure): ‘‘We . . . believe some conceptual, political, and normative
attention should be given to the possibility of state nations. The states we would
like to call state nations are multicultural, and sometimes even have significant
multinational components, which nonetheless still manage to engender strong
identification and loyalty from their citizens, an identification and loyalty that
proponents of homogeneous nation states perceive that only nation states can
engender.’’ They went on to say that neither Switzerland nor India were, in the
French sense, ‘‘strictly speaking a nation state, but we believe both can now be
called state nations. Under Jawaharlal Nehru, India made significant gains in
managing multinational tensions through skillful and consensual usage of numer-
ous consociational practices. Through this process India became in the 1950s and
the early 1960s a democratic state nation.’’3
Nation-state policies stand for a political-institutional approach that attempts
to match the political boundaries of the state with the presumed cultural bounda-
ries of the nation, or vice versa. Needless to say, the cultural boundaries are far
from obvious in most cases; thus, the creation of nation-state involves privileging
one sociocultural identity over other potential or actual sociocultural cleavages
that can be politically mobilized. Nation-state policies have been pursued histori-
cally by following a variety of routes, from relatively soft to downright brutal: (1) by
creating or arousing a special kind of allegiance or common cultural identity
among those living in a state; (2) by encouraging the voluntary assimilation of
those who do not share that initial allegiance or cultural identity into the nation
state’s identity; (3) by using various forms of social pressure and coercion to
achieve this and to prevent the emergence of alternative cultural identities or to
erode them, should they exist; and (4) by resorting to coercion that might, in the
extreme, involve ethnic cleansing.
By contrast, state-nation policies stand for a political-institutional approach
that respects and protects multiple but complementary sociocultural identities.
State-nation policies recognize the legitimate public and even political expression
of active sociocultural cleavages, and they include mechanisms to accommodate
competing or conflicting claims made on behalf of those divisions without impos-
ing or privileging, in a discriminatory way, any one claim. State-nation policies
3. See the chapter titled ‘‘Stateness, Nationalism, and Democratization,’’ in Linz and Stepan,
Problems of Democratic Transition and Consolidation, esp. p. 34 and figure 2.1.
comparative theory and political practice 5
involve crafting a sense of belonging (or ‘‘we-feeling’’) with respect to the state-
wide political community, while simultaneously creating institutional safeguards
for respecting and protecting politically salient sociocultural diversities. The ‘‘we-
feeling’’ may take the form of defining a tradition, history, and shared culture in an
inclusive manner, with attachment to common symbols of the state, or of in-
culcating some form of ‘‘constitutional patriotism.’’
In democratic societies, the institutional safeguards constitutive of state-nation
policies most likely take the form of federalism, often specifically asymmetrical
federalism, or consociational practices.4 Virtually every longstanding and rela-
tively peaceful contemporary democracy in the world whose polity has more than
one territorially concentrated, politically mobilized, linguistic-cultural group that
is a majority in some significant part of the territory is not only federal but also
‘‘asymmetrically federal’’ (Belgium, Canada, and India).5 This means that, up to a
certain point, these polities, in order to ‘‘hold together’’ their great diversity in one
democratic system, had to embed in the constitution special cultural and his-
torical prerogatives for some of the member units, prerogatives that respond to
their somewhat different linguistic or cultural aspirations, demands, and historical
identities.
We believe that had political leaders in India, Belgium, Spain, and Canada
attempted to impose one language and culture on their countries and insisted on
imposing a homogenizing nineteenth-century French-style unitary nation-state
model, the causes of social peace, inclusionary democracy, and individual rights
would not have been served in any of these four longstanding democratic states.
This was so because more than one territorially based linguistic-cultural cleavage
had already been activated in each of these four countries. The strategic question,
therefore, was whether to attempt to repress or accommodate this preexisting politi-
cally activated diversity.
We will also argue that the application of majoritarian nation-state policies
after 1956 in the then-peaceful multicultural democracy of Ceylon eventually
contributed directly to the civil war that beset the country from the early 1980s to
4. We accept Robert Dahl’s definition of federalism as ‘‘a system in which some matters are
exclusively within the competence of certain local units—cantons, states, provinces—and are consti-
tutionally beyond the scope of the authority of the national government; and where certain other
matters are constitutionally outside the scope of the authority of the smaller units.’’ See ‘‘Federalism
and the Democratic Process,’’ in his Democracy, Liberty, and Equality (Oslo: Norwegian University
Press, 1986), p. 114.
5. Sri Lanka, from independence in 1948 until the civil war about Tamil independence which
began in the early 1980s, was an exception. The United Kingdom is a multinational society with
Scottish, Irish, and Welsh assemblies, but English is spoken by the vast majority of the population in
all three areas.
6 crafting state-nations
2009. Further, we will argue that if Burma or China ever were to attempt a
democratic transition, it would be very useful for them to have other models in
mind than just the nation-state model. For example, Ukraine correctly considered
some variance of the full nation-state model, especially concerning the imposi-
tion of Ukrainian as the only language throughout the state. As another example,
Indonesia is exploring the role of a ‘‘federacy’’ in Aceh, of the sort used in Den-
mark in relation to Greenland and the Faroe Islands, and in Finland to the Åland
Islands. In short, one of our major political and theoretical efforts in this book is to
expand our imaginations beyond the nation-state model and to document that, in
all the above cases, the state-nation model was, and can be, extremely useful for
political practice and for the theory of comparative politics.
Asymmetrical federalism historically emerged in Belgium, Spain, Canada,
and India as a policy response aimed at accommodation. We therefore think that,
as a normative concept, an institutional framework, and a set of historical experi-
ences, ‘‘asymmetrical federalism’’ should be strongly considered, by theoreticians
and political leaders alike, as a possible approach to democracy in polities such as
Burma that have at least two territorially based and politically activated linguistic-
cultural cleavages within the existing state.6 At the same time, federalism is nei-
ther sufficient nor necessary for the establishment of a state-nation. The creation
and maintenance of a state-nation requires a number of diversity-sustaining mea-
sures that are not exhausted by federal instruments. For the same reason, it is
possible for a unitary state, where diversities are not geographically concentrated,
to institute many multicultural practices that may lead to a pattern closer to a
state-nation than a nation-state.7
In sum, then, the idea of the nation associated with the nation-state approach
implies creating one common culture within the state, while the idea of the
nation associated with the state-nation approach can contain more than one
politically-salient culture but nonetheless encourages and requires respect for the
6. On ‘‘asymmetrical federalism,’’ see Klaus von Beyme, ‘‘Die Asymmetrisierung des Post-
modernen Föderalismus,’’ in Die Reformierbarkeit der Demokratie: Innovationen und Blockaden, ed.
Renate Mayntz and Wolfgang Streeck (Frankfurt/Main: Campus, 2003), pp. 239–258; and Rainer
Bauböck, ‘‘United in Misunderstanding: Asymmetry in Multinational Federations,’’ ICE Working
Paper Series, Austrian Academy of Sciences, No. 26 (May 2002), Vienna, Austria.
7. For example, Luxembourg is a unitary state but a state-nation. Ukraine, even though it is a
unitary state, is following many state-nation policies because most political elites, both ethnic
Ukrainian and ethnic Russian, were worried that strong ethnic Ukrainian nation-state policies
would generate conflicts with the Russophone population, especially in the eastern regions border-
ing Russia, where 90% of the Ukrainian citizens speak Russian. Also, given the above and irredentist
sentiments in Russia, a policy of decentralization but not necessarily federalization might be pru-
dent. See chapter 6 in this volume.
comparative theory and political practice 7
common institutions of the state and for existing sociocultural diversities. The
analytical distinction between nation-state and state-nation, as the terms imply,
involves an affinity—since both include the term nation; certainly, for some theo-
rists of nationalism, both terms would fit under their conception of a nation.
There is also of course an affinity of both terms with the great importance of
the state.
Thus, state-nation is a term introduced to distinguish democratic states that do
not, and can not, fit well into the classic French-style nation-state model based on
a ‘‘we-feeling’’ resulting from an existing or forged homogeneity. For the differ-
ence between nation-state and state-nation as ideal types, see table 1.1.
Our advocacy of the term state-nation is also based on our recognition that in
some countries, cultural groups are not territorially concentrated but instead are
so diffusely located that even ‘‘asymmetrical federalism’’ is not an option. How-
ever, given the robustness of these different politically salient cultural groups, a
classic French-style nation-state may not, in our times, be an option for a peace-
ful democracy without a costly and most likely nondemocratic period of state-
imposed assimilation efforts and possibly even ethnic cleansing. Nonetheless, in
the same cultural context, a state-nation may prove to be the most viable demo-
cratic model to pursue.
Our introduction of the term state-nation is intended both to establish a nor-
mative standard to which multinational democracies can aspire and to introduce
a set of observable empirical sociopolitical realities that a polity, if it is a state-
nation, will manifest.
Although nation-state and state-nation at one level are analytic ideal-type dis-
tinctions, they can be operationalized using a range of indicators. State-nations
can manage, and have managed, to create powerful and positive citizen identifica-
tion with the institutions and symbols of the state, such as a constitution, inclusive
democratic institutions and procedures, and guarantees of basic freedoms.
We expect a diverse polity, if it has become a state-nation, to have the follow-
ing four empirically verifiable patterns. First, despite multiple cultural identities
among the citizens of the polity, there will be at the same time a high degree of
positive identification with the state and pride in being citizens of that state.
Second, citizens of the state will have multiple but complementary political identi-
ties and loyalties. Third, there will be a high degree of institutional trust in the
most important constitutional, legal, and administrative components of the state.
Fourth, by world democratic standards, there will be a comparatively high degree
of positive support for democracy across all of the diverse groups of citizens in the
country as well as for the specific state-wide democratic institutions through
8 crafting state-nations
table 1.1
Two Ideal Types of Democratic States: ‘‘Nation-State’’ and ‘‘State-Nation’’
Nation-state State-nation
Preexisting condi- Awareness of, and attachment to, one Awareness of, and attachment to, more
tions: Sense of major cultural civilizational tradition. than one cultural civilizational tradi-
belonging/‘‘we- With minor exceptions, this cultural tion within the existing boundaries.
feeling’’ identity corresponds to existing state However, these attachments do not
boundaries. preclude identification with a common
state.
State policy: Cul- Homogenizing attempts to foster one Recognition and support of more than
tural policies core cultural identity, particularly one one cultural identity (particularly rec-
official language; non-recognition of ognition of more than one official lan-
multiplicity of cultures. Unity in one- guage), even more than one cultural
ness. nation, all within a frame of some com-
mon polity-wide symbols. Unity in di-
versity.
Institutions: Ter- Unitary states or symmetrical federa- Federal system. Often de jure or de
ritorial division of tions. facto asymmetrical. Can even be a uni-
power tary state if aggressive nation-state pol-
icies are not pursued and de facto state
multilingual areas are accepted.
Federacies possible.
Politics
Ethno-cultural Not very salient. Salient and recognized and democrat-
or territorial ically managed.
cleavages
Autonomist Autonomist parties are normally not Autonomist parties can govern in fed-
and/or secessio- ‘‘coalitionable.’’ Secessionist parties are eral units and are ‘‘coalitionable’’ at the
nist parties outlawed or marginalized in demo- center. Nonviolent secessionist parties
cratic electoral politics. can sometimes participate in the demo-
cratic political process.
Citizen orientation
Political identity Single identity as citizens of the state Multiple but complementary identi-
and overwhelmingly as members of the ties.
same cultural nation.
Obedience/loy- Obedience to the state and loyalty to Obedience to the state and identifica-
alty the nation. tion with institutions, neither based on
a single national identity.
Multinational
Political Models and Institutional Structures
A
Pure
Democracy
in One State
Is Improbable.
State-Nation
A Democratic
Nation-State Is
Difficult, but
State-Nation
A Democratic Possible.
Nation-State
Is Possible;
Nation-State
State-Nation
Not Necessary.
Low High
Intensity of Political Activation of Multiple, Territorially Concentrated,
Sociocultural “National” Identities
that the polity is more porous, and has a substantially lesser degree of ‘‘stateness,’’
than either a nation-state or a state-nation.8 The numerous small circles and ovals
within this space indicate that there is really a cluster of emergent nation-states
within this weak state. The situation depicted in the upper right-hand corner is
inherently unstable as a single democratic state. The two most likely reequilibra-
tions are: (1) the aspirant nation-states become independent states and the pre-
vious single state fragments or is reduced to a rump; or (2) there is an attempt at
authoritarian re-centralization by a major ethno-political military component of
the threatened state.
In the late 1980s the ethno-federal states of the former U.S.S.R. and Yugoslavia
could analytically be said to have occupied space in the upper-right corner of
figure 1.1. At last count, twenty-five near–nation-states have emerged, often with
substantial bloodshed and repression, out of these two ethno-federal states. But is
it right to assert, as many have done, that all ethno-federal arrangements are ‘‘state-
subverting’’?
8. For a discussion of the concept of ‘‘stateness,’’ see Linz and Stepan, Problems of Democratic
Transition and Consolidation, ch. 2. See also the seminal article by J. P. Nettl, ‘‘The State as a
Conceptual Variable,’’ World Politics 20 (July 1968), pp. 559–592.
comparative theory and political practice 11
supporting the idea of a common state as possible. They may advocate a state that
is a sum of nations, each with its own exclusive identity, symbols, and laws, in
which the state becomes an empty shell and the citizens of the state have nothing
important to say about common institutions—except to the extent that in inter-
national relations and organizations, the states and their citizens have a say.
It is crucial to understand the differences between a state-nation, even when it
contains important multinational components, and a ‘‘pure’’ multinational state
composed of nations, in which the state would be reduced almost to an ‘‘empty
shell,’’ or at the most a confederation, rather than a federation.
In a relatively early formulation (unpublished, but dating from 2000), Linz
argued: ‘‘It is difficult to define the difference between a multinational state and a
state nation. It could be argued that any stable multinational state would also be a
state nation since it requires some sense of identification, of loyalty, to the state
rather than wishing its disintegration.’’9 Elsewhere in the same essay, Linz also
stressed that the kind of affective attachment that the state-nation inspires in
its citizens is something that cannot be fully captured by Habermas’s overly ratio-
nalistic conception of ‘‘constitutional patriotism’’ alone. Linz argued: ‘‘Although
the Verfassungspatriotismus—the commitment to be a liberal-democratic-social
constitutionalism—is one of the elements of the legitimation of state nations, I do
not believe that it is the only or a sufficient one. The construction of a state nation
requires other elements of symbolic and emotional nature we still do not know
well.’’10
Since this earlier formulation, we have become more acquainted with a series
of theoretical arguments for ‘‘multinational federalism’’ that have recently pro-
liferated among Catalan nationalist intellectuals in the Spanish context (per-
haps most prominently exemplified by the work of Ferran Requejo and Miquel
Caminal).11 What strikes us about these accounts is that, despite the language
9. Juan J. Linz, ‘‘Democratic States, Nation States, State Nations, and Multinational States,’’
unpublished. Quotation from pp. 12–13.
10. Ibid., p. 35.
11. See, for example, Miquel Caminal, El federalismo pluralista: del federalismo nacional al
federalismo plurinacional (Barcelona: Paidós, 2002); and Ferran Requejo, Federalisme, per a què?
(Valencia: L’hora del present, 1998). Quite strikingly, a section of the Indian left has often advocated
a multinational state along similar lines. See, for example, Prakash Karat, ‘‘Theoretical Aspects of
the National Question’’ and Irfan Habib, ‘‘Emergence of Nationalities,’’ both in a special edition
of Social Scientist titled ‘‘The National Question in India,’’ no. 37 (August 1975). Karat is now
the general secretary of the Communist Party of India (Marxist) [CPI(M)], India’s largest Commu-
nist party. However, in practice, the Indian Communist parties no longer support the hollowing out
of the state by different ‘‘nationalities.’’
comparative theory and political practice 13
they use, they seem to be arguing not for a truly federal state at all but rather for a
model of confederation between multiple nation-states. In particular, their hos-
tility to the Habermasian conception of ‘‘constitutional patriotism’’ and even to
the related concept of Bundestreue (roughly, ‘‘loyalty to the federation’’)—for
allegedly being unfriendly to diversity itself—alarms us. For our sense is that
commitment to Habermas’s conception of ‘‘constitutional patriotism’’ alone can-
not provide a sufficient basis for any state to be perceived as legitimate—that
something more is necessary; yet, to our dismay, we have found that these other
theorists of ‘‘multinational federalism’’ object to Habermas’s conception on the
opposite grounds. For them, a state that demands ‘‘constitutional patriotism’’ from
all of its citizens is demanding too much. This has led us to suspect that these
intellectuals are in fact using the conceptualization of ‘‘multinational federalism’’
as a mere tactic for legitimating the hollowing-out of the state altogether, for
chipping away at its sovereignty, even as they attempt to legitimate projects of
piecemeal nation-state building at the periphery. Hence the shift in emphasis in
this chapter from the position taken earlier by Linz. We think it urgent to under-
score the necessity for some form of basic loyalty to the institutions of the state,
some form of symbolic attachment to it and identification with it, some form of
‘‘we-feeling’’ even though it makes the project of demonstrating the utility of state-
nations more difficult.
That said, we need to keep in mind the fact that the concept of state-nation,
like that of the nation-state, is an ideal type that only imperfectly corresponds to
any given empirical case. We also need to keep in mind the fluidity of empirical
reality—that is, that cases that once fit closer to the model of the nation-state can
and have evolved into something closer to state-nations (witness Spain and the
UK) and vice versa (witness Austria). With these caveats already in mind, we have
formulated our concept of the state-nation and its partial distinction from both the
nation-state and the ‘‘pure’’ multinational state.
Again, though, these are analytical distinctions that can be operationalized
using a range of indicators. What is more, the distinctions themselves are fluid;
some theorists of nationalism would include in their definitions of nation-states
our state-nations. This, however, would obscure crucial differentiating facts and
have the dangerous political implication of creating demands on citizens based
on the ideals of a nation-state.
Multinationalism is more a sociological conceptualization than a particular
type of political institutionalization. Multinational societies cannot be nation-
states (in a specific sense of the term) but can be either state-nations or the basis
14 crafting state-nations
12. Here we differ substantially with Ernest Gellner. On this point, see Alfred Stepan, ‘‘Modern
Multinational Democracies: Transcending a Gellnerian Oxymoron,’’ in his Arguing Comparative
Politics (Oxford: Oxford University Press, 2001), pp. 181–199.
comparative theory and political practice 15
is crucial, multinational and multicultural societies can be, or can become, state-
nations with much ‘‘we-feeling,’’ if complementary as well as multiple cultural and
political identities exist, are nurtured, or even are generated.
Another issue concerns the relation between multicultural societies and the
competing conceptions of the nation-state and the state-nation. Our conception
of state-nation derives from our belief, based on historical case studies and analy-
sis, that democracy is possible in polities that are sociologically and politically
multicultural and even partly (but not exclusively) multinational, if an effort is
made to legitimate the state by those minorities and majorities who could con-
ceivably aim at its disintegration.
At this point, we need to address what many may see as a powerful argument
from the opposite end against our idea of a state-nation. After the bloody dis-
integration of Yugoslavia and parts of the Soviet Union, many analysts have begun
to reject wholesale all political and institutional frameworks that grant any form
of prerogatives to territorially concentrated sociocultural groups—arrangements
they call ‘‘ethno-federal’’ or ‘‘national federal.’’ These scholars criticize ‘‘ethno-
federal’’ arrangements because they believe that such arrangements privilege
‘‘subnational’’ sociocultural identities at the expense of identification with com-
mon symbols, institutions, and individual rights. This privileging, they claim, is
likely to foster the activation of conflictual, as opposed to complementary, identi-
ties and, in some cases, violence and fragmentation.13
These critics ignore, however, that nearly all successful democratic states with
more than one politically activated, territorially concentrated, linguistic-cultural
majority have institutional frameworks that include a substantial (but absolutely
13. Even though in her book, Subversive Institutions: The Design and Destruction of Socialism
and the State (New York: Cambridge University Press, 1999), Valerie Bunce does not explicitly argue
this point, many scholars who read her book have employed its analysis of the Yugoslavian and
Soviet experiences to make the case that ‘‘ethno-federal’’ institutions in and of themselves are
‘‘subversive’’ institutions for stateness and peace. See for example David John Meyer, ‘‘Ethnic
Territorial Autonomy and Post-Soviet Ethnic Political Mobilization,’’ unpublished Ph.D. disserta-
tion, Department of Political Science, Columbia University, 2007. For a related argument, see Jack
Snyder, From Voting to Violence: Democratization and Nationalist Conflict (New York: Norton,
2000). For her part, Bunce refers to ‘‘national federalism’’ as one of the most important ‘‘subversive
institutions.’’ She argues that national federalism helped ‘‘produce over time a ‘disintegration’ of the
Soviet, Yugoslav, and Czechoslovak states along republican lines’’ and that ‘‘with the expanded
opportunities for major change in the 1980s, ‘disintegration’ in all three instances translated into
actual disintegration, and the state and the regime departed from Europe in virtual tandem’’ (p. 102).
Likewise, she maintains that ‘‘national federalism worked to build nations (or reinforce such pro-
cesses, if nations were already formed), along with protostates at the republican level’’ (p. 136) and
that this ‘‘contributed not just to homogeneity within republics, but also to diversity among re-
publics, and that it was the latter that made a single-state project untenable, especially in turbulent
times’’ (p. 140).
16 crafting state-nations
become state-nations. This means, quite simply, that for consolidated democracy
to be viable, these states would have to craft institutional frameworks that contain
both (a) a substantial ethno-federal dimension and (b) mechanisms facilitating
identification with common symbols and institutions. If, in the process of democ-
ratization, leaders of these states were to pursue either a pure nation-state model
or a pure ethno-federal model of the sort we label ‘‘pure multinationalism’’ in
figure 1.1, the result would almost certainly be continued armed struggle and
failure to achieve democratic consolidation.
Let us not allow a reading of the Yugoslavian and Soviet experiences to destroy
the legitimacy of all institutional arrangements containing an ethno-federal di-
mension; to do so would require giving up on the middle ground of the state-
nation—a model that has proven valuable in the important but extremely difficult
task of reconciling cultural inclusiveness with democratic stability in states con-
taining more than one politically activated, territorially concentrated ‘‘national’’
sociocultural group.
However, if we argue that ethno-federal arrangements in the state-nation ideal
type are not necessarily state-subverting mechanisms, it is incumbent upon us to
explicate a well-argued alternative. What are the identifications, norms, practices,
and institutions that can facilitate the construction of a democratic polity close to
a state-nation ideal type, even in a politically robust multinational society? This
takes us to the interlinked set of policies that help craft successful ‘‘state-nations.’’
On theoretical and empirical grounds we would like to make the case that there
are arrangements that cohere in an unusual, almost counterintuitive, ‘‘nested
policy grammar’’ that may facilitate the emergence and persistence of a state-
nation.15 By nested, we mean that each of the seven policies mentioned below is
most likely to be implemented, or facilitated, if the previous policy in the se-
quence has already been adopted. Seven phrases are an intrinsic part of this
grammar:
15. The Oxford English Dictionary defines grammar as a ‘‘means of indicating the relations of
words.’’
18 crafting state-nations
Again, these policies are ‘‘nested’’: the second policy, ‘‘group recognition,’’ is
normally nested within the first, federalism (especially asymmetrical federalism);
the fourth policy, involving the coalitionability of what we call centric-regional
parties, is greatly facilitated if the choice of the third policy has been parliamen-
tarianism because the executive is a sharable good; and successful achievement of
complementary identities depends heavily on the success of the previous six
policies.
1. Why an asymmetrical federal state? A federal state, rather than a unitary state,
is part of the nested grammar of a state-nation because federal state structures
allow a large territorially concentrated cultural group with serious nationalist
aspirations and possibly a language with its own script to attain self-governance
within that territory. Why asymmetrical? In a symmetrical federal system, all units
must have identical rights and obligations. It is politically possible, however, that
some territorially concentrated and culturally diverse groups have in their history
acquired prerogatives that they desire to retain or reacquire, and it is also possible
that some tribal groups that control a large territory (such as the Mizos in India)
would only agree to remain in or join the federation if some of their laws pertain-
ing to such matters as land use or education, found nowhere else in the polity,
were respected. Bargains and compromises on these issues, which might be neces-
sary for peace, and voluntary membership in the political community are nego-
tiable in an asymmetrical system but are normally unacceptable in a symmetrical
system.
2. Why both individual rights and collective recognition? The polity would not
be democratic unless throughout the polity individual rights are constitutionally
inviolable and state-protected. This necessary function of the center cannot be
devolved. But in Charles Taylor’s sense, some territorially concentrated cultural
groups, even nations, may need some collective recognition for rights beyond
classic liberal rights (or what Michael Walzer calls ‘‘Liberalism 2’’) for members of
some groups to thrive culturally or even possibly to exercise fully their classic
individual liberal rights.16 Walzer argues that Liberalism 2 ‘‘allows for a state
16. See Charles Taylor, ‘‘The Politics of Recognition,’’ in Multiculturalism, ed. Amy Gutmann
(Princeton: Princeton University Press, 1994). An elegant development of a variant of this argument
in found in Joseph Raz, The Morality of Freedom (Oxford: Oxford University Press, 1986), esp. chs. 8
comparative theory and political practice 19
and 10, and his Ethics in the Public Domain: Essays in the Morality of Law and Politics (Oxford:
Clarendon Press, 1994), esp. preface and chs. 1, 6, and 8.
17. The quotation from Michael Walzer is from his ‘‘Comment’’ in Multiculturalism, ed. Gutt-
man, p. 99. For a somewhat different approach to group recognition, see Will Kymlicka’s discussion
of ‘‘group specific rights’’ in his Multicultural Citizenship: A Liberal Theory of Minority Rights
(Oxford: Clarendon Press, 1995), esp. ch. 4.
18. See Alfred Stepan, ‘‘The World’s Religious Systems and Democracy: Crafting the ‘Twin
Tolerations,’ ’’ in his Arguing Comparative Politics, pp. 213–254. Also see Alfred Stepan, ‘‘The Multi-
ple Secularisms of Modern Democratic and Non-Democratic Regimes,’’ in Rethinking Secularism,
ed. Mark Juergensmeyer, Craig Calhoun, and Jonathan Van Antwerpen (New York: Oxford Univer-
sity Press, forthcoming). For an authoritative analysis of India’s pioneering ‘‘principled distance’’
form of secularism, see Rajeev Bhargava, ‘‘What is Secularism For?’’ in Secularism and Its Critics,
ed. Bhargava (Delhi and Oxford: Oxford University Press, 2004), pp. 486–542. Also see his ‘‘Political
Secularism’’ in The Oxford Handbook of Political Theory, ed. John S. Dryzek, Bonnie Honig, and
Anne Phillips (Oxford: Oxford University Press, 2006), pp. 636–655.
20 crafting state-nations
and business. Furthermore, if citizens can pursue such public and private polity-
wide careers, the incentives to ‘‘exit’’ from these polity-wide networks will be
weaker.19
5. Why politically integrated but not culturally assimilated? In a state-nation,
many cultural and especially ethno-national groups will be educated and self-
governing in their own language. They will thus probably never be fully assimi-
lated to the dominant culture in the polity. Indeed, any attempt to assimilate these
groups would invite resentment, resistance, and perhaps rejection of the system.
This is a reality of state-nations, which distinguishes them from nation-states,
where cultural assimilation is a possibility and very often a reality.
The absence of cultural assimilation does not preclude the possibility of politi-
cal integration. If the ethno-federal group sees the polity-wide state as having
helped put a ‘‘roof of rights’’ over its head, and if the ‘‘centric-regional’’ parties are
‘‘coalitionable’’ with polity-wide parties and regularly help form government at
the center and many individuals from the ethno-federal group also participate in,
and feel they benefit from, polity-wide careers, then they are politically integra-
table into the polity-wide state-nation.
6. Why cultural nationalists versus secessionist nationalists? Ernest Gellner
forcefully articulated the position of many nation-state theorists when he famously
asserted, ‘‘Nationalism is primarily a political principle, which holds that the
political and the national unit should be congruent. . . . Nationalist sentiment
is the feeling of anger aroused by the violation of the principle. . . . A nationalist
movement is one actuated by a sentiment of this kind.’’20 Similarly, we are con-
stantly admonished not to advocate state-nation policies because all national-
ism inevitably becomes ‘‘secessionist nationalism’’ with eventual demands for
independence.
However, we can have a situation in which a ‘‘cultural-nationalist’’ movement,
nested within asymmetrical federal and parliamentary systems, wins democratic
political control of a component unit of a federation; governs and educates the
citizens of its territory in the language, culture, and history of its nation; and is also
coalitionable at the center. If such a group is challenged by secessionist national-
ists who use, or threaten to use, violence in order to become independent, the
ruling nationalist group would risk losing the treasured resources they have ac-
quired. Given that, it is entirely possible that the cultural nationalists would use
19. However, the systematic effort by an ethno-cultural group to monopolize access to careers,
even in their ethno-federal unit, runs counter to the nurturing and preservation of a state-nation.
20. All quotations are from the influential opening paragraphs of Ernest Gellner’s Nations and
Nationalism (Oxford: Oxford University Press, 1983), p. 1, emphasis in original.
22 crafting state-nations
the political and security resources now under their control against the seces-
sionist nationalists.21
7. Why not only multiple identities but also complementary identities? In the
non-zero-sum polity-wide system produced by the six nested policies and norms
we have just discussed, it is likely that citizens in the multinational society would
strongly identify with, and remain loyal to, both their culturally powerful ethno-
federal unit and the polity-wide center. Most citizens would have such comple-
mentary identities because the center has recognized and defended many of their
cultural demands and, in addition, helped structure and protect their full par-
ticipation in the overall politics of the polity. Such citizens are also likely to have
strong trust in the center because they see the center, and the institutions histor-
ically associated with it, as helping to deliver some valued collective goods such as
independence from a colonial power, security from threatening neighbors, and
participation in a large common market. Thus the pattern of complementary and
multiple identities that is likely to obtain is not just an accident; this outcome
would have been earned through the deliberate crafting of policies.
21. However, it cannot be denied that if there is a conflict between secessionist nationalists and
the central state apparatus over the use of force, there might be a convergence on some issues in the
dispute between cultural nationalists and the secessionist nationalists.
comparative theory and political practice 23
transition to a federal type of state after the death of Franco, asserted that Spain is a
state for all Spanish citizens, a nation-state for a large part of the population, and
only a state but not a nation for important minorities. Also, he added, there is a
small minority that contests or rejects that state and seeks independence.22
In the Spanish case, like that of quite a few other countries, would-be nation-
builders who sought to create a unique shared sense of identity based on language,
history, and culture following the French model ultimately failed. We would
argue that such efforts in the twentieth century were often not fully successful; in
the twenty-first century, they might well backfire and arouse the latent sense of
national identity of significant minorities.23 In 1993, in a paper titled ‘‘State Build-
ing and Nation Building,’’ Linz cogently formulated some of the main reasons
why we feel this is the case. He stressed in particular how in today’s world,
sensibilities have emerged within the ‘‘international community’’ that act as an
effective pressure against it:
22. Juan J. Linz, ‘‘Early State-Building and Late Peripheral Nationalisms against the State: The
Case of Spain,’’ in Building States and Nations: Analyses by Region, vol. 2, ed. S. N. Eisenstadt and
Stein Rokkan (Beverly Hills, CA: Sage Publications, 1973), pp. 32–116.
23. In India the term minority normally refers only to a religious minority. However, in this book
we follow standard social science vocabulary when we use the word minority to include linguistic,
tribal, ethnic, and religious minorities.
24. Linz, ‘‘State Building and Nation Building.’’
24 crafting state-nations
Given the significant technological changes that have occurred since the late
nineteenth century state-induced homogenization processes so well described
by Eugen Weber, and the analytically distinct but related emergence of what
Charles Taylor calls ‘‘the politics of recognition,’’ there are grounds for thinking
such processes are now less available. Most of the world’s minorities can keep in
cultural contact with their home cultures via radio, cassettes, and cheap air
travel to a vastly greater extent than was possible a hundred years ago. Also, due
to advances in literacy and communications, more minority communities have
semiprofessional ‘‘cultural carriers,’’ in the Weberian sense of Träger, than a
hundred years ago. Normative changes in the form of increased desire for cul-
tural autonomy in some minority (especially Muslim) communities—contested
by rising antiforeign sentiments in the majority cultures that reduces the inte-
grating capacity that in theory the majority culture would like—probably have
contributed to greater cultural will, and greater cultural capacity, for minorities
to resist cultural assimilation.25
cerned with their rights as a culture. This makes the assimilating policies of
the successful nineteenth-century nation-builders, aimed at erasing such distinct
identities, as in France, extremely costly. However, paradoxically, the minority
that has gained control of ethno-cultural political institutions can often pursue
nation building or exclusionary policies that large states—the center—cannot
pursue without being sanctioned.
In our view, as we shall show in chapter 2, India at certain times has been and
is, like Spain, Belgium, and Canada, to mention three democratic federal states
with multinational components, a nation-state for most citizens, a state to which
they owe allegiance but not a nation in the classical sense for significant minori-
ties, and a state that is contested by some minorities in the periphery of the state.26
This brings us to a basic distinction among federal states between those federa-
tions that are largely ‘‘coming together’’ in their origin versus those that are largely
‘‘holding together’’ in their origin.27 ‘‘Coming together’’ federations are formed by
a process in which relatively autonomous separate units, often sharing much of
the same political culture, sometimes a common enemy, jointly arrive at an
agreement to pool part of their previous sovereignty in order to gain the advan-
tages of creating a new federal state. This was the case of the United States, when
the thirteen colonies got together to achieve a more perfect union as an indepen-
dent state. The history of Australia fits that same pattern, as does even multi-
cultural Switzerland in the course of its long history. This is the standard theory of
how federal states are created.28
But there is another quite different process for the emergence of new federal
states. Old states, governed as unitary states and originally conceived as future
nation-states, when confronted with rising peripheral nationalisms, with new
identities based on language, culture, or history that threaten their unity, can turn
26. For a typology of democratic states that takes into account both whether they are unitary or
federal ones and whether they are mononational or multinational, see Juan J. Linz, ‘‘Para un mapa
conceptual de las democracias,’’ Politeia, no. 26 (2001), pp. 25–46.
27. A more detailed discussion of the necessary distinction between ‘‘coming together’’ federal-
ism and ‘‘holding together’’ federalism can be found in Alfred Stepan, ‘‘Toward a New Comparative
Politics of Federalism, (Multi)Nationalism, and Democracy: Beyond Rikerian Federalism,’’ in his
Arguing Comparative Politics, pp. 315–361.
28. The classic statement of what we call ‘‘coming together’’ federations is found in William
Riker, The Development of American Federalism (Boston: Kluwer Academic Publishers, 1987),
pp. 17–42. See also his ‘‘Federalism’’ in Handbook of Political Science, vol. 5, ed. Fred Greenstein
and Nelson W. Polsby (Reading, MA: Addison-Wesley, 1975), pp. 93–172. Riker implies that autono-
mous political units who come to believe that their security and economy will be enhanced if they
create a federation often tend to do so. However, we want to note that no country, once it was an
established nation-state, has ever created a ‘‘coming together’’ federation. Certainly, on security
grounds in the interwar years, the Baltic states and the Scandinavian nation-states had significant
reasons to create federations but did not.
26 crafting state-nations
29. It is possible that if the princely states in India had coincided with cultural, linguistic, and
other social characteristics and the federation had been created, as was sometimes discussed in the
1930s by the princely states retaining their identities and acceding to a federation, India could have
been a case of a ‘‘coming together’’ federation. For many reasons we shall not discuss here, this did
not happen. For the atmosphere of the debate in the 1930s, see N. D. Varadachariar, Indian States in
the Federation (Calcutta: Oxford University Press, 1936).
30. In the case of the Soviet Union, especially in 1919 to 1923, there was actually a third pattern
that Stepan calls ‘‘putting together’’ federalism. See his ‘‘Russian Federalism in Comparative Per-
spective,’’ Post-Soviet Affairs, no. 16 (April–June 2000), pp. 133–176.
31. For an overview of the constitutional impasse in Canada, see Richard Simeon, ‘‘Canada:
Federalism, Language and Regional Conflict’’ in Federalism and Territorial Cleavages, ed. Ugo M.
Amoretti and Nancy Bermeo (Baltimore: Johns Hopkins University Press, 2004), pp. 93–122; and
Simeon, ‘‘Debating Secession Peacefully and Democratically: The Case of Canada,’’ in Democ-
racies in Danger, ed. Alfred Stepan (Baltimore: Johns Hopkins University Press, 2009), pp. 41–56.
On the evolution of Québécois nationalism, see Maurice Pinard, ‘‘Les quatre phases du mouve-
ment indépedantiste québécois,’’ in Combat Inachevé, ed. Robert Bernier, Vincent Lemieux, and
Maurice Pinard (Sainte-Foy, Quebec: Presses de l’Université du Quebec, 1997).
comparative theory and political practice 27
32. For the original formulation of the concept of ‘‘constitutional patriotism,’’ see Dolf Stern-
berger, Verfassungspatriotismus (Frankfurt am Main: Insel-Verlag, 1990). For Jürgen Habermas’s
development of the concept, see his Einbeziehung des Anderen: Studien zur politschen Theorie
(Frankfurt am Main: Suhrkamp, 1996). For a recent elaboration on the theme (in English) and
adaptation of it to contexts outside of Germany, see Habermas, ‘‘Citizenship and National Identity,’’
in Between Facts and Norms (Cambridge, MA: MIT Press, 1998), pp. 491–515.
33. On the recent emergence of national consciousness in Austria, see T. Bluhm, Building an
Austrian Nation: The Political Integration of a Western State (New Haven: Yale University Press,
1973), esp. pp. 220–241. On the same theme, see also the excellent piece by Fritz Plasser and Peter A.
Ulram, ‘‘Politisch-Kulturell Wandel in Österreich,’’ in Staatsbürger oder Untertanen? Politische
Kultur Deutschlands, Österreichs und Schweiz im Vergleich, ed. Plasser and Ulram (New York:
P. Lang, 1991), pp. 157–245.
28 crafting state-nations
we call a state-nation, where the institutions of the state, with its distinctive political
culture, are the basis of a particular type of identification of its citizens.34
In contrast to the ideal-type state-nation of Switzerland, in state-nations with
important multinational components, such as Spain, Belgium, or Canada, many
citizens, who may constitute a significant proportion of the population of federal
units, identify with a distinctive territorially based culture that some influential
members of the community see as a nation, with its own language, history, rights,
and grievances against the state in which they live. The federal state-nation is a
nightmare to those who originally conceived of the state as a nation-state; a
nightmare to those who want to nationalize the whole population in the process of
nation building, of which the French Republic would be the historically most
successful model; a nightmare to those nation builders for whom federalism
would be conceived, at the most, as a form of decentralization for purposes of
administrative efficiency. It is also a nightmare to nationalists who want to create a
separate nation-state.
There are those who think that the multinational federal state is inevitably con-
demned to break up, who see federalism in those states as only a step toward
disintegration and who therefore want to limit the federal constitution and engage
in a process of more or less aggressive nation building. For complex reasons we
have already noted, such efforts are likely to fail, producing a backlash that will
lead to the opposite result from the one that their proponents pursue. However,
intelligent political engineering, constructive political leadership, and some fa-
vorable contextual factors can serve to overcome the tension inherent in multi-
national societies. A federal state that is multinational can become a successful
state-nation. Unfortunately, we have few systematic studies of how this has been
34. A fine overview of the Swiss case can be found in Lidija Basta, ‘‘Minority and Legitimacy of
a Federal State,’’ in Federalism and Multiethnic States: The Case of Switzerland, ed. L. Basta and
Thomas Fleiner (Fribourg, Switzerland: Institute of Federalism, 1996), pp. 41–69. For another
treatment that deals extensively with language policy in Switzerland, see Kenneth D. McRae,
Conflict and Compromise in Multilingual Societies, vol. 1 (Ontario, Canada: Wilfred Laurier Uni-
versity Press, 1986). For an earlier formulation of the concept of the state-nation, and its distinctive-
ness from both the nation-state and the multinational state, see Juan J. Linz, ‘‘Democratic States,
Nation States, State Nations, and Multinational States.’’ A shortened version of this article was
published in German as ‘‘Nationalstaaten, Staatsnationen und multinationale Staaten,’’ in Staat,
Nation, Demokratie. Traditionen und Perspektiven moderner Gesellschaften. Festschrift für Hans-
Jürgen Puhle, ed. Marcus Gräser, Christian Lammert, and Söhnke Schreyer (Göttigen, Germany:
Vandenhoeck und Ruprecht, 2001), 27–38. See also Juan J. Linz, ‘‘Democracia, multinacionalismo
y federalismo,’’ in Revista Española de Ciencia Política 1 (October 1999), pp. 7–40.
comparative theory and political practice 29
achieved. That is why we believe that our study of the Republic of India and its
history and institutions can make an important contribution to this important task
for social scientists and policy-makers. Before we turn to a detailed study of the
crafting of a state-nation in India in chapter 2, we first take on the suspicion that
multinational federations cannot draw positive identification from its citizens and
then get them to posit trust in its institutions. We consider some evidence from
Spain and Belgium with regard to multiple but complimentary identities before
turning to a comparative dataset for the level of citizen trust in some of the key
institutions in eleven federal democracies.
It might seem natural for our task to draw on evidence and justification from
the proponents of multiculturalism. Unfortunately, the brilliant theorizing about
multiculturalism of recent years, particularly in the United States and, to a lesser
extent, Europe, is only partly relevant to this task.35 Multiculturalism in the way
that we find it discussed in that literature is not distinctive to federal states. The
literature is equally relevant to unitary states like France, with its increasingly
important Muslim immigrant population.36 The literature on multiculturalism
is especially relevant to cultural minorities, particularly immigrants claiming a
range of rights as individuals and communities without rising to the level of
territorially based autochthonous communities with an articulated or latent na-
tional identity.
Multiculturalism represents a different dimension of social and political reality
that we can find in nation-states, state-nations, and multinational societies. Also,
multiculturalism certainly can be found in India as a whole and within the states
of the Indian federation.
Most of the literature on nationalism treats national identities as if they were
mutually exclusive. The literature is plagued with the use of expressions like ‘‘the
35. The literature on multiculturalism is of course extensive, and we refer here only to some of
the most basic works, written from a variety of perspectives. These include: Brian Barry, Culture and
Equality: An Egalitarian Critique of Multiculturalism (Cambridge, MA: Polity Press, 2001); Seyla
Benhabib, The Claims of Culture: Equality and Diversity in the Global Era (Princeton: Princeton
University Press, 2002); Will Kymlicka, Liberalism, Community, and Culture (New York: Oxford
University Press, 1989) as well as his more recent contribution, Multicultural Citizenship: A Liberal
Theory of Minority Rights (New York: Clarendon Press, 1995); Bikhu Parekh, Rethinking Multi-
culturalism: Cultural Diversity and Political Theory (Basingstoke, UK: Macmillan, 2000); Taylor,
‘‘Politics of Recognition’’; James Tully, Strange Multiplicity: Constitutionalism in an Age of Diversity
(New York: Cambridge University Press, 1995); and Iris Marion Young, Justice and the Politics of
Difference (Princeton: Princeton University Press, 1990).
36. There is a growing selection of high-quality literature on precisely this point. See, for
example, Ahmet Kuru, ‘‘Secularism, State Policies, and Muslims in Europe: Analyzing French
Exceptionalism,’’ Comparative Politics 4 (2008), pp. 1–19; and John R. Bowen, Why the French Don’t
Like Headscarves: Islam, the State, and Public Space (Princeton: Princeton University Press, 2006).
30 crafting state-nations
Catalans’’ or ‘‘the Flemish’’ and their opposites, ‘‘the Spanish’’ and ‘‘the Belgians.’’
However, such expressions are a gross oversimplification. Though nationalists on
both sides reject the idea of dual identities as a form of bigamy, in fact, in all more
or less multinational societies, most citizens tend to have dual but often comple-
mentary, or at least not exclusive, identities.
Identities in Spain
The region of Catalonia, in Spain, provides a case in point. Since the late nine-
teenth century and particularly in the twentieth century, there has been a growing
interest in cultural and, increasingly, national identities among people in bi-
lingual regions in certain parts of Spain—most acutely, in Catalonia and the
Basque Country.37 At the turn of the century, nationalist parties emerged in both
of these regions and began to articulate these identities. In the decades following
the Spanish Civil War, such identities gained strength as a reaction to the Franco
regime, since that regime pursued an aggressive policy against peripheral na-
tionalist movements, including discriminatory language policies. By the end of
the Franco era, the democratic opposition had come to sympathize with the
peripheral nationalist movements and even to demand that their aspirations be at
least partly recognized. In the transition to democracy, the drafters of the 1978
constitution did just that; they agreed to accommodate linguistic, cultural, and
national differences by organizing the state as an ‘‘estado de autonomías,’’ a type of
federal political system.38
37. For a useful bibliography on the historiographical debates about the rise of peripheral
nationalisms in Spain, see Xosé-M. Núñez, ‘‘Historical Research on Regionalism and Peripheral
Nationalism in Spain: A Reappraisal,’’ published as a working paper by the European University
Institute in Florence as ECS no. 92/6 (1992). For Linz’s contribution to this debate, see Juan J. Linz,
‘‘Early State-Building and Late Peripheral Nationalisms against the State: The Case of Spain,’’ in
Building States and Nations, vol. 2, ed. Eisenstadt and Rokkan, pp. 32–116.
38. On the process of devolution to a federal state in Spain, see Juan J. Linz, ‘‘Spanish Democ-
racy and the Estado de las Autonomías,’’ in Forging Unity Out of Diversity, ed. Robert A. Goldwin,
Art Kaufman, and William A. Schambra (Washington, DC: American Enterprise Institute for
Public Policy Research, 1989), pp. 260–303. On electoral results in and public opinion about the
Estado de las Autonomías during the first fifteen years of democracy, see Juan J. Linz, ‘‘De la crisis de
un estado unitario al Estado de las Autonomías,’’ in La España de las Autonomías, ed. Fernando
Fernández Rodríguez (Madrid: Instituto de Estudios de Administración Local, 1985), pp. 527–672.
See also Linz for the five alternative questions on national identity. On the continuing conflict in the
Basque Country, see Juan J. Linz, Conflicto en Euskadi (Madrid: Espasa Calpe, 1986). Also see
Francisco J. Llera, Los Vascos y la Política. El proceso político vasco: elecciones, partidos, opinión
pública y legitimación en el País Vasco, 1977–1992 (Bilbao: Servicio Editorial Universidad del País
Vasco, 1994). On public opinion in the Basque Country, see Euskalherria en la encuesta Europea de
valores (Bilbao: Universidad de Deusto, 1992). Also see the series of Euskobarometro, directed by
Francisco Llera (Bilbao: Servicio Editorial Universidad del País Vasco). For public opinion in
comparative theory and political practice 31
Identities in Belgium
Nor is Spain an exception in this regard; for the same is true in the case of
Belgium. Belgium was founded in 1830 as an independent unitary parliamentary
monarchy. In the course of a complex process marked by considerable conflict, it
evolved in the twentieth century into a federal, basically bi-national and bilingual
federal democracy.39 In Belgium, a number of relevant questions about national
identity have been asked in opinion polls too, distinguishing among those who
speak Dutch, those who speak French, and the inhabitants of Brussels. All of these
again reveal the predominance of multiple but complementary identities. For
Catalonia, see the yearly surveys published by the Institut de Ciències Polítiques i Socials (Barce-
lona: Universitat Autònoma de Barcelona); and Francisco Andrés Orizo and Maria-Àngels Roque,
Cataluña 2001: Los catalanes en la encuesta Europea de valores (Madrid: La Fundación Santa
María, 2001). See also Thomas Jeffrey Miley, Nacionalismo y política lingüística: el caso de Cata-
luña (Madrid: CEPC, 2006).
39. For a good synthesis of this historical process and an extended discussion of language policy
there, see McRae, Conflict and Compromise in Multilingual Societies, vol. 1, ch. 1. For a study that
focuses on linguistic conflict in the metropolitan region of Brussels, the only place where significant
numbers of French-speakers and Flemish-speakers live side by side, see Jan de Volder, ‘‘Le FN Brade
Bruxelles,’’ in Revue Française de Geopolitique, no. 6 (May 1998).
32 crafting state-nations
table 1.2
Multiple but Complementary Identity in Spain:
Subjective National Identity in Spain, Catalonia, and Basque Country
Basque
Identity All of Spain Catalonia Country
Only Spanish 13 9 6
More Spanish than Cat/Basque/other 8 4 6
As Spanish as Cat/Basque/other 56 41 34
More Cat/Basque/other than Spanish 14 27 23
Only Cat/Basque/other 6 16 27
Do not know/no answer 4 3 4
(N) (10,476) (1,200) (1,800)
Sources: For the Basque Country, Catalonia, and Galicia, Sondeo de opinión del Observatorio
Político Autonómico: 2003 (Barcelona: ICPS, 2004); for the rest of Spain and all of Spain, CIS
study no. 2455 (2002). These five alternative questions were designed by Linz in 1979 and have
been used in many surveys in Spain and by other scholars.
Note: All figures in column percentages may not add up to 100 due to rounding. N, number
of respondents in each column.
example, in Flanders, which many analysts see as moving towards secession, the
following question was asked: ‘‘Which of the following statements applies most to
you: ‘I consider myself only as a Belgian,’ ‘I feel more Belgian than Flemish,’ ‘I
feel as Belgian as Flemish,’ ‘I feel more Flemish than Belgian,’ or ‘I feel only
Flemish?’ ’’ What is most surprising about the poll is that only 4% of the Flemish
population self-identified as ‘‘only Flemish.’’ Indeed, in 1995 the modal form of
self-identification in Flanders was the multiple but complementary identity ‘‘as
Belgian as Flemish.’’40
Nonetheless, many analysts still predicted that an ‘‘only Flanders’’ sense of
identity would grow steadily and imperil the unity of Belgium. What trends in
identity, if any, seem to be occurring in Belgium? In surveys, the three-year
average for 1980, 1981, and 1982 of Flemish respondents who said their strongest
‘‘feeling of belonging’’ was with Flanders was 47%. However, far from increasing,
the three-year average for the same question for the years 1997, 1999, and 2003 had
dropped twenty points, to 27%.41 Furthermore, younger respondents were more
40. The data we use for Belgium in this survey are based on the 1995 General Election Study,
conducted by the Interuniversitair Politieke-Opinieonderzoek, K. U. Leuven, and the Point d’appui
Interuniversitaire sur l’Opinion publique et la Politique, U. C. Louvain, results published in 1998.
41. See Jaak Billiet, Bart Maddens, and André-Paul Frognier, ‘‘Does Belgium (Still) Exist?
Differences in Political Culture between Flemings and Walloons,’’ West European Politics 29, no. 5
(2006), pp. 912–932.
comparative theory and political practice 33
table 1.3
Multiple but Complementary Identity in Belgium:
Subjective National Identity by Region
All of
Identity Belgium Flanders Wallonia Brussels
Only Belgian 14 11 18 24
More Belgian than Flemish/Walloon 21 17 25 27
As Belgian as Flemish/Walloon 43 45 44 32
More Flemish/Walloon than Belgian 17 23 10 11
Only Flemish/Walloon 3 4 2 3
Do not know/no answer 2 1 2 5
(N) (3,651) (2,099) (1,258) (311)
Source: 1995 Belgian General Election Study. See footnote 40.
Note: All figures in column percentages have been rounded off. N, number of respondents in
each column.
inclined to identify with Belgium than older respondents. In 1999 only 12% of
Flemish respondents supported Flemish independence.42
Among the French-speaking Walloons, who at the time of the founding of the
state lived in the more prosperous state-building community, identification with
the Belgian nation is somewhat stronger than it is for the whole of the population:
18% of them identify themselves as ‘‘only Belgian’’; 25% identify themselves as
‘‘more Belgian than Walloon’’; and 44% consider themselves ‘‘as Walloon as
Belgian.’’ A mere 10% identify themselves as ‘‘more Walloon than Belgian’’; and a
miniscule 2% feel ‘‘only Walloon.’’ In the capital city of Brussels, the only place in
the country where significant numbers of French-speakers and Dutch-speakers
live side by side, 24% identify themselves as ‘‘only Belgian’’; 26% identify them-
selves as ‘‘more Belgian than Flemish/Walloon’’; 31% report an equal dual iden-
tity; 11% identify themselves as ‘‘more Flemish/Walloon than Belgian’’; and only
3% fail to mention the Belgian identity. In fact, only 15% of the population of
Brussels speak Flemish. So a movement for an independent Flanders would run
the risk of weakening or severing valued ties with Brussels for Flemish-speakers,
because the citizens of Brussels would certainly not vote, given the chance, to
leave Belgium to join a new state of Flanders (see table 1.3).
What’s more, not only do the overwhelming majority of citizens in Belgium,
regardless of the territory from which they hail, identify themselves at least some-
42. Wilfred Swenden and Marten Theo Jans, ‘‘Will It Stay or Will It Go? Federalism and the
Sustainability of Belgium,’’ West European Politics 29, no. 5 (2006), pp. 877–894.
34 crafting state-nations
times as Belgians, they also register a very high degree of affective attachment to
an important common state institution—specifically, the monarchy. Such attach-
ment is evident in the responses of Belgian citizens to a question about how much
they trust their king—for fully 54% of them claim to trust him either ‘‘very much’’
or ‘‘a lot’’ (13% and 41%, respectively); while a mere 11% of them claim to trust him
‘‘only a little’’ or ‘‘very little’’ (6% and 5%, respectively).43
We understand this kind of affective attachment to a set of common institu-
tions, symbols, and places such as Brussels to be indispensable for the legitimacy,
and therefore stability, of any state with a high level of cultural, linguistic, and
even national heterogeneity. This is why we stress the importance of not only
multiple but complementary identities within a multinational, federal, demo-
cratic framework. Of course, as we have already suggested, there are two inti-
mately related difficulties with this framework: first, that centralists often dream of
doing away with the fact of multiple identities; and second, that peripheral na-
tionalists often seek to undermine the fact of complementary identities. But, at
least in the Belgian case, neither of these difficulties seems to be unmanageable.
Belgium is certainly adopting more and more confederal policies, which many
argue are decreasing overall state efficiency, but we do not share the skepticism of
some other commentators, who feel that Belgium is falling apart.44 Both the
overwhelming preponderance of dual identities and, especially, the high level of
affective attachment to common symbols and institutions justify our sense of
optimism. Were such affective attachment to common symbols and institutions
lacking, there would be reason for pessimism. Late in the history of Yugoslavia as a
federal state, for example, it is highly doubtful that any Yugoslavian institution had
a high level of trust shared by most of the citizens of the country. Fortunately,
however, the Belgian case is quite different from that of Yugoslavia.
Yugoslavia’s ‘‘pure multinationalism,’’ as Bunce has shown, led to the hollow-
ing out of almost all authoritative state-wide functions and to increasingly polar-
ized and conflictual identities. In sharp contrast, Belgium’s ‘‘holding together’’
federalization and other state-nation-building processes from 1970 to 1993 did
neither. Liesbet Hooghe, in her detailed article on Belgium’s ‘‘holding together’’
federalization since 1970, brilliantly documents what has happened, and what has
not, to identities, ethnic conflicts, and central state functions.
Concerning identities, Hooghe argues that ‘‘from the mid-1980s, Belgian iden-
tity has gained ground on regional and local identities. The development of
regional governance institutions in Belgium has not gone hand in hand with a
deepening of regional identity. The data show, instead, that a system of multilevel
governance may encourage the development of complementary multiple identi-
ties. . . . The survey results also confirm that, like in Spain and the European
Union (EU), regional and national identities are not exclusive, but complemen-
tary.’’ Analyzing Walloon-Flemish conflicts and protests, Hooghe concludes that
‘‘disruptive, non-violent territorial protest became widespread in the 1950s and
1960s. It topped in the late 1970s, and it has declined since. Through much of the
postwar period Flemish and Walloon identities gained strength at the expense of a
Belgian identity, but in the 1990s Belgian and regional identities appear to have
become more inclusive. The ebbing of disruptive, territorial protest and of exclu-
sive identities coincides with the transformation from a unitary to a federal state in
1993.’’ Regarding the polity-wide functions of the new ‘‘holding together’’ federal
state, Hooghe correctly notes that in the process of moving from a unitary to a
federal state, Belgium saw a significant transfer of functions to the sub-units; she
even uses the phrase ‘‘Hollowing the Center’’ as her subtitle. However, in great
contrast to Yugoslavia, many of the core attributes found in democratic federal
states in the world still remain under the control of the Belgian federal center. As
Hooghe stresses, ‘‘The list of exclusive federal competencies is . . . substantial:
defense, justice, security, social security, fiscal and monetary policy.’’ She also
notes that central ‘‘federal institutions remain the prime venue for the resolution
of much horizontal Flemish-Francophone conflict, and here the familiar con-
sociational mechanisms are unchanged.’’45
Thus, in terms of our analytic categories, Belgium’s often overlooked reintegra-
tion since the 1970s has come about largely because it has adopted ethno-conflict–
ameliorating state-nation policies, such as ‘‘holding together’’ federalism. In com-
plete contrast, from 1970 to 1991 Yugoslavia increasingly adopted ethno-conflict–
intensifying, disintegrative ‘‘pure multinational’’ policies.
By all means, nationalists would like the question formulated not as, ‘‘Are you
more Flemish than Belgian?’’ or ‘‘Are you more Catalan than Spanish?’’ but rather
as, ‘‘Are you either Flemish or Belgian?’’ or ‘‘Are you either Catalan or Spanish?’’—
despite the ubiquity of multiple but complementary identities in settings that
are more or less multinational. And inevitably, both those who speak of self-
45. Liesbet Hooghe, ‘‘Belgium: Hollowing the Center,’’ in Federalism and Territorial Cleavages,
ed. Ugo M. Amoretti and Nancy Bermeo (Baltimore: Johns Hopkins University Press, 2004), pp. 55,
65, 74.
36 crafting state-nations
determination, that is, the right of every nation to become an independent state,
and those who favor total national integration into a single cultural or linguistic
community reject the very idea of dual identities.
This is the main reason (and there are many) why democratic plebiscites are
normally such an undesirable solution. People have to make one choice or an-
other, like the one of defining the territorial units for which the decision should be
binding. The quorum necessary to reverse such a decision is totally different from
a normal election since it cannot be reversed four or five years hence. A plebiscite
might be the only solution in certain extreme situations where the polarization
created by violent conflict has destroyed any dual identity. But in those cases, it
will mean a loss of rights and equal citizenship among those not supporting the
majoritarian choice.
The state-nation concept would seem to have some utility for advanced de-
mocracies and the democratic management of a modest number of language
groups. But what is the ‘‘scope value’’ of the concept? Is it appropriate in much
poorer countries, in countries with much greater diversity? For this we turn to
India, the poorest, the most diverse (and the most populous) democracy of long
standing in the world in the next chapter.
We would like to conclude with a cautionary note, however. The attitudes
documented in table 1.3 indeed support a state-nation, but we must acknowledge
that we do not rule out political leaders acting in ways not congruent with the
attitudes of their constituents and constantly pushing for autonomist or secession-
ist policies. We also do not rule out that a majority of people in Brussels and
Wallonia might end up wishing that Flanders would exit Belgium. However, the
reality is that, despite tensions, it is not clear that a majority within Flanders would
be eager to embrace a political formula where they would politically and econom-
ically lose Brussels, the de facto capital of Europe.
A major claim of nation-state theorists and advocates is that only a nation-state can
generate the necessary degree of trust in the major institutions of the state that a
modern democracy needs to function well.
Let us do a simple empirical test of this claim by examining comparative trust
in the entire universe of federal systems that have been democratic for at least
twenty-five continuous years. In our judgment, there are eleven such countries; in
alphabetical order, they are Australia, Argentina, Austria, Belgium, Brazil, Can-
ada, Germany, India, Spain, Switzerland, and the United States.
comparative theory and political practice 37
Fortunately, the World Values Survey was carried out in all eleven of the coun-
tries.46 We thus have data for the degree of trust for six key political institutions for
all eleven countries. These key institutions are: the central government, the legisla-
ture, the legal system, the civil service, political parties, and the police. All of the
data are presented in table 2.9, but we are interested here in a summary measure.
To proceed with our test, let us now divide our eleven countries into those
countries that fit the nation-state ideal type in that they are (1) symmetrically
federal; (2) have no constitutionally embedded ethno-federal dimensions; and
(3) are not multilingual as a matter of law. These countries are Austria, Germany,
Australia, the United States, Brazil, and Argentina.
Those countries in our country set that fit the state-nation ideal type in that
they are (1) asymmetrically federal; (2) have constitutionally embedded ethno-
federal features; and (3) are constitutionally multilingual are Belgium, Canada,
Spain, and India. We will add Switzerland to this set because, even though it is
symmetrically federal, Switzerland is much closer to a state-nation type than a
nation-state type, based on the attributes detailed in table 1.1.
When we examine the average country trust scores for the six key political
institutions for each of these polities, we get very surprising results (see table 1.4).
The ‘‘closer to the nation-state, the greater the trust’’ claim is obviously not
supported by these data. If we look at the average ranking of trust (the lower the
number the greater the trust), state-nations receive an average ranking of 4.8,
whereas nation-states score much worse, with an average ranking of 7.0. The
conclusion remains the same if we replace the crude measure of average ranking
for a more precise average of percentages, as state-nations record an average of
47% trust in their institutions, whereas nation-states have an average of 39% trust.
The same pattern obtains even if we throw out Argentina, which scores signifi-
cantly lower on trust than the other countries and is something of an outlier. The
nation-state set still has a trust average of 43%, four points lower than the state-
nation set.47
The claim that ‘‘any use of ethno-federal devices is subversive of the state’’ is also
obviously not supported by these data because all of the countries close to the state-
nation pole have some ethno-federal features as well as strong polity-wide trust.
46. Five waves of this survey have been done since it began in 1981. These five waves have been
carried out in 97 societies. The World Values Survey is under the overall supervision of Ronald Ingle-
hart, the program director for the Institute of Social Research at the University of Michigan. For a
discussion of these surveys, see appendix A in Pippa Norris and Ronald Inglehart, Sacred and Secular:
Religion and Politics Worldwide (Cambridge: Cambridge University Press, 2004), pp. 243–246.
47. All figures computed from the data in table 2.9.
38 crafting state-nations
table 1.4
State-Nations More Trusted than Nation-States:
Average Trust Ranking in State Institutions in Longstanding Federal Democracies
Average trust ranking
(the lower the number,
the greater the trust)
States closest to fitting state-nation model
India 3.0
Switzerland 3.7
Canada 4.0
Belgium 6.3
Spain 7.2
States closest to fitting nation-state model
Brazil 4.6
Austria 5.0
United States 5.9
Germany 7.6
Australia 8.0
Argentina 10.6
State-nation average 4.8
Nation-state average 7.0
Source: World Values Survey. For a more extensive discussion of these sources, see the lengthy
footnote to table 2.9.
Note: Average trust ranking was arrived at by ranking the world’s eleven longstanding federal
democracies by level of citizen trust in each of six institutions (the country with the highest level
of trust in that institution was ranked 1, and so on) and then averaging the six ranks for each
country. The data from which the ranks were derived is reported in table 2.9.
48. For disaggregated data on pride for all eleven countries, see table 2.2.
chapter two
India as a State-Nation
Shared Political Community amidst
Deep Cultural Diversity
India would appear to be one of the most difficult cases for our argument that
multiple but complementary identities and democratic state-nation loyalties are
possible even in a polity with significant ‘‘robust multinational’’ dimensions as
well as intense linguistic and religious differences. For many of its citizens, India is
a nation-state; for others, it is what we call a state-nation. However, as we discussed
in chapter 1, India also has some unmistakable dimensions of a multinational
society. Think of the popular, enduring sentiment for ‘‘freedom’’ in the Kashmir
Valley, the short-lived militant Sikh-led Khalistan movement in the Punjab, the
longer armed rebellion in Mizoram and Nagaland, and the potentially secession-
ist Dravidian movement in the south, among others.1 In addition, Indians have
1. Chapters 3 and 4 contain extensive analysis and documentation of each of these cases of
‘‘robust multinationalism’’ involving some separatist movements. At this point, let us simply say, if
we divide Jammu and Kashmir into its three regions (Kashmir, Ladakh, and Jammu), the Kashmir
region meets our definition of being ‘‘robustly multinational.’’ The Kashmir Valley is a territorially
concentrated linguistic-cultural majority of Muslims who speak Kashmiri. Also, significant armed
groups with overt and covert support from Pakistan and varying degrees of support from the people
in Kashmir Valley have waged a battle for ‘‘Azadi,’’ or freedom. For the conflicts concerning the
national and international status of Kashmir and its independence, see Perspectives on Kashmir: The
Roots of the Conflict in South Asia, ed. Raju G. C. Thomas (Boulder, CO: Westview Press, 1992);
and Sumantra Bose, The Challenge in Kashmir (London: Sage, 1997). On independence move-
ments and secessionist wars in the Northeast, see Sanjoy Hazarika, Strangers of the Mist: Tales of
War and Peace from India’s Northeast (London: Penguin, 1994); and Ved Marwah, Uncivil Wars:
Pathology of Terrorism in India (New Delhi: Harper Collins, 1995). On the Khalistan Independence
Movement in the Punjab, see Surinder S. Jodhka, ‘‘Looking Back at the Khalistan Movement: Some
Recent Researches on its Rise and Decline,’’ Economic and Political Weekly 36, no. 16, April 21–27,
2001, pp. 1311–1318. On short-lived pro-Dravidistan separatist movements, see Eugene Irschick,
Politics and Social Conflict in South India: The Non-Brahmin Movement and Tamil Separatism,
40 crafting state-nations
had to nurture, defend, and deepen democracy in a social and political context in
which this multinational dimension interacts with more linguistic and religious
diversity, and greater poverty, than found in any other longstanding democracy in
the world. While all of this makes our task difficult, these difficulties enhance the
scope value of India for building an analytical as well as a normative case for state-
nation. Let us first attend to some of the complexities and specificities of India.
1916–1929 (Berkeley and Los Angeles: University of California Press, 1969); and Narendra Subrama-
nian, Ethnicity and Populist Mobilization: Political Parties, Citizens, and Democracy in Southern
India (Oxford: Oxford University Press, 1999).
2. For an analytic discussion of these figures, see Jyotirindra Das Gupta, Language Conflict and
National Development: Group Politics and National Language Policy in India (Berkeley: University
of California Press, 1970), pp. 31–68.
3. Ibid.
4. Census of India, 2001, Office of The Registrar General of Census, Statement 6, available at
http:/ /censusindia.gov.in/CensuseDatae2001/CensuseDataeOnline/Language/Statement6.htm,
accessed March 29, 2010.
5. His oft-cited judgment about the impossibility of having more than one important function-
ing language and significant nationality in a democratic polity was ‘‘free institutions are next to
impossible in a country made up of different nationalities. Among a people without fellow-feelings,
especially if they read and speak different languages, the united public opinion necessary to the
working of representative institutions cannot exist.’’ See Mill, Considerations on Representative
Government (1861), in Utilitarianism, On Liberty, Considerations on Representative Government, ed.
Geraint Williams (London: Everyman, 1993), p. 393.
india as a state-nation 41
parativists because it has been developed in the context of great religious diversity.
Indian society has large communities of almost every major world religion—
Hindu, Islamic, Buddhist, Sikh, and Christian. Even after Partition in 1947, India
had a large Islamic population. In 2009, India’s Muslim population constituted a
‘‘minority’’ of approximately 161 million people, which made it the third-largest
Islamic population in any country in the world, exceeded only by Indonesia’s 203
million and Pakistan’s 174 million.6 At a time when many scholars and political
activists worry about the ‘‘clash of civilizations,’’ and some see Islamic society as
being in deep cultural conflict with democracy, we find it very useful to docu-
ment and analyze that the world’s largest Islamic community with extensive dem-
ocratic experience is in multicultural, multinational, federal, and consociational
India. India pioneered in the democratic world a form of secularism that gave
‘‘equal support and equal respect’’ to all of India’s religions; in part due to this
formula, citizens from all religions have shown a very high degree of support for
Indian democracy and trust in the Indian state.7
Conceptually and comparatively, India’s poverty raises important intellectual
challenges, especially for those who might posit that state-nation norms are a lux-
ury reserved for wealthy countries. One of the most enduring propositions in
social science is Seymour Martin Lipset’s formulation that democracy correlates
very strongly with overall socioeconomic development.8 Arend Lijphart did not
6. Pew Forum on Religion in Public Life, ‘‘Mapping the Global Muslim Population: A Report
on the Size and Distribution of the World’s Muslim Population,’’ October 2009, p. 14, available at
www.pewforum.org.
7. Numerous tables based on various national representative surveys by CSDS to follow in
chapters 2, 3, and 4 document this assertion. An essential book on India’s model of secularism, with
important contributions by Rajeev Bhargava, Amartya Sen, Akeel Bilgrami, Charles Taylor, and
Ashis Nandy, is Rajeev Bhargava, ed., Secularism and Its Critics (Delhi and Oxford: Oxford Univer-
sity Press, 2004). For Muslims, see the important work by Mushirul Hasan, Legacy of a Divided
Nation: India’s Muslims since Independence (Delhi: Oxford University Press, 1997). For Hindu
fundamentalism, the standard text is Christophe Jaffrelot, The Hindu Nationalist Movement in
India (London: Hurst, 1996). On India’s model of secularism in comparative perspective, see Alfred
Stepan, ‘‘The Multiple Secularisms of Modern Democratic and Non-Democratic Regimes,’’ in
Rethinking Secularism, ed. Mark Juergensmeyer, Craig Calhoun, and Jonathan Van Antwerpen
(New York: Oxford University Press, forthcoming).
8. The classic initial formulation of this argument is Seymour Martin Lipset, ‘‘Some Social
Requisites of Democracy: Economic Development and Political Legitimacy,’’ American Political
Science Review 53 (March 1959), pp. 69–105. Larry Diamond reviewed three decades of literature
relevant to the development/democracy debate and concluded that the evidence broadly supports
the Lipset proposition; see Diamond, ‘‘Economic Development and Democracy Reconsidered,’’ in
Re-Examining Democracy, ed. Gary Marks and Larry Diamond (Newbury Park, CA: Sage, 1992), pp.
93–139. Linz and Stepan discuss Lipset’s proposition in their Problems of Democratic Transition and
Consolidation: Southern Europe, South America, and Post-Communist Europe (Baltimore: Johns
Hopkins University Press, 1996), p. 77. Yadav put together evidence to show that socioeconomic
42 crafting state-nations
status and electoral participation do not correlate positively in India. See his ‘‘Understanding the
Second Democratic Upsurge: Trends of Bahujan Participation in Electoral Politics in the 1990s,’’ in
Transforming India: Social and Political Dynamics of Democracy, ed. Francine R. Frankel, Zoya
Hasan, Rajeeva Bhargava, and Balveer Arora (Delhi: Oxford University Press, 2000), pp. 120–145.
9. For the list of the twenty-one countries, see Arend Lijphart’s classic Democracies: Patterns of
Majoritarian and Consensus Government in Twenty-One Countries (New Haven: Yale University
Press, 1984), p. 38.
10. Arend Lijphart, Patterns of Democracy: Government Forms and Performance in Thirty-Six
Countries (New Haven: Yale University Press, 1999), table 4.1. Also see Lijphart’s article, ‘‘Demo-
cratic Institutions and Ethnic/Religious Pluralism: Can India and the United States Learn from
Each Other—and from the Smaller Democracies?’’ in Democracy and Diversity: India and the
American Experience, ed. K. Shankar Bajpai (Oxford: Oxford University Press, 2007), pp. 14–49.
11. Calculated using GDP per capita in 1985 at Purchasing Power Parity (PPP) and current
international prices. India’s GDP per capita in that year was US$919, in extremely sharp contrast to
average GDP per capita of the other twenty countries (US$13,093). Data are from World Bank,
World Development Indicators: 2006 (Washington, DC: International Bank for Reconstruction and
Development/World Bank, 2006).
12. All data from World Bank, World Development Indicators: 2008 (Washington, DC: Interna-
tional Bank for Reconstruction and Development/World Bank, 2008). Figures are given in GDP per
capita using current international purchasing power parity dollars (PPP).
india as a state-nation 43
world.13 The next most populous democracy is the United States, with a popula-
tion of 300 million. The combined population in 2006 of the only other long-
standing multinational federal democracies—Spain, Canada, and Belgium—was
less than 85 million, less than half the population of Uttar Pradesh, the most
populous of India’s twenty-eight states.
A final reason why the study of Indian democracy should be of particular
interest to comparativists is precisely the exclusion of the Indian experience, or
even Indian scholars, from the mainstream democratization literature. This ab-
sence has impoverished comparative politics and, more important, our imagina-
tions of politically possible alternatives for democracies.14
Given its extraordinary diversity, the project of creating a democratic nation-
13. India is set to surpass China as the most populous country in the world by 2040, according to
the 2004 revision of World Population Prospects by the Population Division of the United Nations.
14. The founding literature in democratization studies neglects India nearly entirely. Because of
its focus on Latin America and southern Europe, there are no chapters on India in the pioneering
four-volume series on democratization edited by Guillermo O’Donnell, Philippe Schmitter, and
Laurence Whitehead, Transitions from Authoritarian Rule (Baltimore: Johns Hopkins University
Press, 1986). Because of Linz and Stepan’s focus on postcommunist Europe, southern Europe, and
Latin America, they too never discuss India. There are, of course, many excellent books on Indian
politics with important implications for democratization theory. Rajni Kothari’s classic Politics in
India (Boston: Little Brown, 1970) remains almost untapped for its larger implications. The only
partial exception to this neglect was Myron Weiner, a comparativist by instinct and training, who
wrote several essays spelling out the comparative insights from his work on India. See Gabriel A.
Almond, ‘‘Myron Weiner on India and the Theory of Democratization,’’ in India and the Politics of
Developing Countries: Essays in Memory of Myron Weiner, ed. Ashutosh Varshney (New Delhi:
Sage, 2004). The distinguished Indian scholar of the politics of language, Jyotirindra Das Gupta,
wrote the chapter on India in the three-volume work edited by Larry Diamond, Juan J. Linz, and
Seymour Martin Lipset, Democracy in Developing Countries (Boulder, CO: Lynne Rienner, 1989).
In the main, Indian scholars’ best-known work on politics has been on such themes as caste,
nationalism, colonialism, and on the politics of the ‘‘subalterns’’ rather than on the interaction
between political processes and democratic institutions. Three excellent critical reviews and bibli-
ographies on Indian, and non-Indian, scholarly writings about politics in India can be found in
Partha Chatterjee, ed., State and Politics in India (Delhi: Oxford University Press, 1997), pp. 566–
576; Sudipta Kaviraj, ed., Politics in India (Oxford: Oxford University Press, 1997), pp. 1–36; and
Sunil Khilnani, The Idea of India (London: Hamish Hamilton, 1997), pp. 217–242. For more recent
contributions with wider implications, see, for example, Atul Kohli, ed., The Success of India’s
Democracy (Cambridge: Cambridge University Press, 2001); Pranab Bardhan, The Political Econ-
omy of Development in India (New Delhi: Oxford University Press, expanded edition 1998); Ash-
utosh Varshney, Ethnic Conflict and Civic Life: Hindus and Muslims in India (New Haven: Yale
University Press, 2002); Pradeep Chhibber and Ken Kollman, The Formation of National Party
Systems: Federalism and Party Competition in Britain, Canada, India, and the U.S. (Princeton:
Princeton University Press, 2004); Kanchan Chandra, Why Ethnic Parties Succeed: Patronage and
Ethnic Head Counts in India (Cambridge: Cambridge University Press, 2004); Steven I. Wilkinson,
Votes and Violence: Electoral Competition and Ethnic Riots in India (Cambridge: Cambridge
University Press, 2006); Partha Chatterjee, The Politics of the Governed: Reflections on Popular
Politics in Most of the World (New York: Columbia University Press, 2004); SDSA Team, State of
Democracy in South Asia (Delhi: Oxford University Press, 2008); and Paul Brass, Ethnicity and
Nationalism: Theory and Comparison (London: Sage, 1991).
44 crafting state-nations
india’s specificity
At this stage we need to take on three common, if unstated, beliefs that have
prevented the Indian experience from being treated as appropriate material for
larger generalizations about democracy and diversity. First, there is general un-
ease about the socioeconomic record of democratic governance in India, giving
rise to a suspicion that India’s success on diversity and democracy may be too good
to be true. Second, there is a tendency to view any example of federalism and
diversity through the prism of the political system of the United States; it appears
thus that India is too different to be relevant for comparative theorization. Finally,
from a very different vantage point, India’s success may appear to be too unique to
allow general lessons. In one way or another, these three beliefs see India as a case
that defies generalization. We examine these beliefs and suggest that while great
attention must be paid to India’s specificity—indeed, ‘‘specificity matters’’ is an
important point this book makes about every single country—this does not render
India so special that it rules her out as a source of general insights.
We are very aware of India’s continuing problems with poverty and with low
levels of literacy, nutrition, basic sanitation, Maoist-style violence (by the ‘‘Naxal-
ites,’’ named after Naxalbari, the place where Maoist militancy in India was born),
violation of human rights and periodic communal riots. Table 2.1 makes some of
these comparative problems abundantly clear. We also recognize that these facts
are very relevant to an overall assessment of the quality of democratic governance
15. John Keane’s recent global history of democracy, The Life and Death of Democracy (Lon-
don: Simon and Schuster, 2009) recognizes the significance of the Indian experience to the global
narrative; see pp. 585–647.
india as a state-nation 45
table 2.1
Comparative Indicators of India’s Human and Income Poverty
Average GDP per capita in purchasing power parity (PPP) in 2000 U.S. $20,252
dollars among Arend Lijphart’s universe of the thirty-six continuous
democracies of the world from at least 1977 to 1996
India’s GDP per capita in PPP in 2007 U.S. dollars $2,753
India’s human development index (HDI) ranking among the 173 134th out of 182
countries of the world ranked by the United Nations Development
Programme (UNDP)
India’s HDI ranking among Arend Lijphart’s thirty-six continuous 34th out of 36
democracies
India’s human poverty index (HPI-1) ranking among 135 ranked states 88th out of 135
Adult female literacy rate in India 54.5%
Percentage of underweight children in India at age 5 46.0%
‘‘Great Poverty’’ Level
All India 22.7%
Muslims 31.0%
States with more than 20 deaths due to Naxalite violence in 2008 6 (18%)
Sources: UNDP, Human Development Report 2009: Overcoming Barriers: Human Mobility and
Development (New York: Palgrave Macmillan, 2009), pp. 173, 177, and 181–184; Government
of India, Prime Minister’s High Level Committee (Sachar Committee), ‘‘Social, Economic, and
Educational Status of the Muslim Community of India,’’ November 2006; and Arend Lijphart,
Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries (New
Haven: Yale University Press, 1999)—table 4.1 shows Lijphart’s universe of the thirty-six
countries in the world, including India, that were continuous democracies in his judgment
from at least 1977 to 1996. For information on Naxalite violence in India, see Ministry of Home
Affairs, Government of India, Naxal Management Division, available at
http:/ /mha.nic.in/uniquepage.asp?IDePK=540.
in India and believe that policies and outcomes in these areas should be, and
could be, substantially improved.
However, this book is not about democracy and its links to development; we
focus here only on how democratic political arrangements of the kind that we
analyze below serve as mechanisms for handling societal diversity and potential
conflict. It is quite clear to any student of democracy and democratization or of
nationalism, multinationalism, and diversity and extreme crises of ‘‘stateness,’’ as
in the U.S.S.R. and Yugoslavia, that India began its democratic experiment with
greater diversity than any longstanding democracy that ever existed. It is also
evident to any analyst of survey data about contemporary India that democracy is
increasingly supported by the overwhelming majority of these diverse groups in
India. This pattern is not sufficiently recognized, much less analyzed, by general
readers or even by most specialist scholars, so one of our major tasks in this
46 crafting state-nations
16. The eleven federal countries that have been functioning democracies for at least the last
twenty years are India, the United States, Switzerland, Germany, Austria, Belgium, Spain, Australia,
Brazil, Argentina, and Canada. Belgium’s long transformation from a unitary to a federal state was
only completed in 1993, but it had increasingly functioned as a federal system since the 1970s.
17. All calculations based on the population figures of the Census of India 2001. It should be
noted that while the Census provides the exact breakdown of language and religion for each state, it
does not provide a breakdown combining the two. This has been estimated here by assuming that all
the Urdu-speakers in the Hindi heartland states are Muslims and that the rest of the language
speakers are evenly distributed across religions.
india as a state-nation 47
18. Findings based on a five-country survey reported in SDSA Team, State of Democracy in
South Asia, p. 73.
19. Ibid., p. 263.
48 crafting state-nations
macy and are able to exercise their authority over most of its population and
territory. As we shall document, the overwhelming majority of Indian citizens
respect the Indian state and generally expect it to serve the collective interests of
its citizens.
In analyzing the Indian experience, one cannot overlook the external aspects,
specifically that the exit options were effectively closed. It is true that the struggle
for independence and the democratic institutions created at that time legitimated
a sense of Indian nationhood and a conception of the nation open to its pluralism.
This makes it unlikely that major political forces, parties, and intellectuals would
favor whatever secessionist demands might appear in the periphery (with the
possible exception of Kashmir). Even the two main communist parties, the Com-
munist Party of India and the Communist Party of India (Marxist), have on bal-
ance sided with the project of preserving the boundaries of the Indian nation-
state, even if by selective use of coercive state apparatus.20 This consensus on the
overriding importance of the Indian nation, and on the Indian state-nation, lends
weight to the provisions in the constitution that allow the government to act in
defense of the Republic of India. We must remember that in India, as in many
other federal constitutions, there are provisions to defend the constitution and the
unity of the state, if necessary by coercive means. The awareness of that possibility,
and the actual record of using those resources in Kashmir, Nagaland, Mizoram,
and the Punjab has left those who might question the state in a ‘‘no exit’’ position,
one that forces the search for negotiated compromise within the context of the
federal institutions of the state.21
20. A case in point was the support by the communist parties for the suppression of Khalistani
secessionism in the Punjab. Unlike, say, the communists in Sri Lanka, the Indian communists did
not extend any support in this instance to the demand of separation from the Indian state. The
communist workers were among the prime targets of the pro-Khalistan militants. A veteran leader of
the Communist Party of India, Satyapal Dang, offers a defense of this stance in his Genesis of
Terrorism: Analytical Study of Punjab Terrorists (New Delhi: Patriot Publishers, 1988). A statement
by the Punjab state committee of the Communist Party of India (Marxist), issued a decade after
terrorism had subsided, still ‘‘expressed deep concern over the increasing activities of Sikh funda-
mentalists and extremists in the state and felt that they posed a threat to the hard-earned peace in the
state and also to the unity and integrity of the country.’’ See People’s Democracy, vol. 29, no. 35,
August 28, 2005.
21. Various human rights groups within and outside India have documented the denial of basic
civil rights guaranteed by the Indian constitution and law and the use of brutal repressive measures
by security forces in the conflicts mentioned above. See the various annual reports and other India
related material of Amnesty International at http:/ /web.amnesty.org/library/eng-ind/index. Various
human rights groups within the country have also extensively documented these violations. See the
various publications of the reputable and independent Peoples’ Union for Civil Liberties (PUCL) at
www.pucl.org, especially its report ‘‘Kashmir: A Report to the Nation,’’ 1993, published by PUCL
and CFD. An academic compilation of these reports can be found in A. R. Desai, ed., Violation of
Democratic Rights in India (Bombay: Popular Prakashan, 1986). For some recent reports, see Ram
india as a state-nation 49
Narayan Kumar, Amrik Singh, Ashok Agrwaal, and Jaskaran Kaur, Reduced to Ashes: The Insurgency
and Human Rights in Punjab (New Delhi: South Asia Forum for Human Rights, 2003), available at
www.punjabjustice.org; and Kashmir: An Enquiry into the Healing Touch, a report by AFDR, HRF
and OPDR, Hyderabad, 2003.
22. According to the Census of India 2001, the population of Nagaland was 1.9 million, about
0.19% of the Indian population. Mizoram’s population was 890,000 and accounted for about 0.086%.
Jammu and Kashmir’s population was 10.1 million. With this absolute number, Jammu and Kashmir
had about 0.99% of the total population of the country (of this, 5.7 million, or 0.56% of the Indian
population lived in the Kashmir Valley, the heart of political alienation). Of these states, only Punjab
has a national population share of more than 1%. Census of India-2001, Primary Census Abstract,
Series 1 & 2: Registrar General of India, New Delhi.
50 crafting state-nations
same time, the comparison will show that India and its democratic institutions
enjoy legitimacy among the citizens equal to, if not greater than, the institutions
in other longstanding democratic federal pluralistic societies. It should be noted
that in no political system do all the citizens grant to the state, its institutions and
democratic processes, a unanimous legitimacy or allegiance. It is only the size of
India and its multitudes that make these problems somewhat more significant
than in some other countries, once we translate the proportions into absolute
numbers of citizens.
imagining a state-nation
There is another, deeper reason to turn to India in our attempt to expand the
imagination of how democracies can deal with diversities. An expansion of the
kind advocated in this book cannot merely be a matter of designing or redesign-
ing institutions; an expansion of democratic imagination must deal directly with
ideas, ideals, and images. We need to engage with contestations around the
difficult relationship between ideas of nationhood and the realities of complex
diversities. Arguably, India has been the most fertile nursery of this kind of ideo-
logical contestation. We believe that democratic theory should draw on and re-
construct the rich discursive contestations that have taken place in India in the
last two centuries.
The building of a state-nation in India was not an accident of history or an
afterthought. We suggest that the state-nation model was implicit in the idea of
India forged by modern Indian political thinkers, nurtured by the freedom move-
ment, eventually enshrined in the Indian constitution, sustained by the first gen-
eration of post-Independence leadership, and institutionalized in competitive
politics thereafter. In that sense our state-nation model is best seen as a new
analytic ideal type as well as a theoretical defense of what political thinkers and
practitioners in India have known for the better part of a century.
This is no doubt a tribute to the foresight and creativity of the leaders of India’s
freedom struggle. But it is equally important to acknowledge that this creativity
itself was in a large measure the outcome of these leaders’ reflections on the
historical situation in which the anticolonial struggle found itself. British colonial
rule produced a body of knowledge about India that denied her an identity except
as a colonial subject. Modern Indian political thinkers found themselves pitted
against this indictment by John Strachey, a British administrator and author of a
very influential book entitled India, published in 1888: ‘‘This is the first and the
most essential thing to learn about India—that there is not, and never was an
india as a state-nation 51
India, or even any country of India, possessing, according to European ideas, any
sort of unity, physical, political, social or religious. . . . That men of the Punjab,
Bengal, the North-Western Provinces, and Madras, should ever feel they belong
to one great nation, is impossible.’’23
The colonial subject could respond to this in many ways. They could accept the
characterization and agree that the very idea of Indian nationalism was incoherent;
this line of reasoning led to collaboration with the colonial regime. Alternatively,
they could argue that India was not one but many nations that warranted more than
one nation-state; this line logically required the partition of the country and
acceptance of British arbitration until it could be effected. These two responses did
have some followers in the nineteenth and the twentieth centuries, but most of the
Indian political thinkers began to question this colonial reading of India.
Thus started the search for the unity of India. This search was initially rather
tentative and defensive but gradually turned more confident and even aggressive.
This search took many forms. History writing and historical novels were the most
popular forms of what Sudipta Kaviraj calls ‘‘the imaginary institution of India,’’
but this search also took the form of sociological or philosophical arguments and
political tracts.24
The colonial reading often encouraged simple counterarguments. One anti-
colonial argument was that India was indeed a nation-state in the European sense
of the term, that it was based on a unity of culture, race, religion, and language.
This argument stressing cultural homogeneity tended to draw exclusively on the
‘‘great tradition’’ of Hindu heritage and to exclude the vast resource of non-Hindu
and ‘‘little traditions’’ of India. This line of reasoning was adopted by some lead-
ing Hindu nationalist ideologues and organizations that stood on the fringes
23. John Strachey, India (London: Kegan Paul, 1888), pp. 5–8, cited in Ainslie Embree, India’s
Search for National Identity (Delhi: Chanakya Publications, 1980), p. 17. This was not an isolated
statement but a sharp articulation of a very common view of the empire and its defenders. In making
the above categorical assertion, Strachey seems not to appreciate fully that the United Kingdom,
itself created only in 1707, has not always been ‘‘one great nation’’ but had robust territories and
identities occupied by Scots, Welsh, Irish, and English. See, for example, Linda Colley, Britons:
Forging the Nation, 1707–1837 (New Haven: Yale University Press, 1992).
24. Sudipta Kaviraj, ‘‘The Imaginary Institution of India,’’ Subaltern Studies 7, ed. Partha Chat-
terjee and Gyananendra Pandey (Delhi: Oxford University Press, 1993), pp. 1–39. For a detailed
analysis of the imaginary histories produced in nineteenth-century Bengal, see Sudipta Kaviraj, The
Unhappy Consciousness: Bankimchandra Chattopadhyay and the Formation of Nationalist Dis-
course in India (Delhi: Oxford University Press, 1995), pp. 107–157. For a scholarly critique of the
idea that the unity of India was a gift of the British Empire, see Radha Kumud Mookerji, The
Fundamental Unity of India (1914; repr. New Delhi: Bharatiya Vidya Bhavan/Chronicle Books,
2003). Partha Chatterjee’s Nationalist Thought and the Colonial World: A Derivative Discourse?
(Delhi: Oxford University Press, 1986) explores the complexities and contradictions that character-
ize the nationalist response to colonial constructions.
52 crafting state-nations
25. M. S. Golwalkar, We, or Our Nationhood Defined, original edition 1939, 4th ed. 1947, cited
in Christophe Jaffrelot, ed., Hindu Nationalism: A Reader (Princeton: Princeton University Press,
2007), pp. 98–99. This volume provides a useful introduction and selections to the ideas of the
leaders of Hindu nationalist organizations such as Hindu Mahasabha, RSS, and the BJP.
26. Rabindranath Tagore, Nationalism (1917; repr. London: MacMillan, 1950), pp. 5, 12. For an
elaboration of Tagore’s critique of nationalism, see Ashis Nandy’s iconoclastic reading, Illegitimacy
of Nationalism: Rabindranath Tagore and the Politics of Self (Delhi: Oxford University Press, 1994).
india as a state-nation 53
recognized and supported. But he also felt that the Independence movement was
mobilizing Indian nationalism into a positive force for India and the world: ‘‘Vio-
lent nationalism, otherwise known as imperialism is the curse. Non-violent na-
tionalism is a necessary condition of . . . civilized life.’’ He saw the multinational,
multilinguistic nationalism of the Independence movement as ‘‘India’s contribu-
tion to peace.’’27 In many of his writings, he insisted that Indian nationalism must
be inclusive and not based on any religion or language: ‘‘If the Hindus believe that
India should be peopled only by Hindus they are living in a dream-land. The
Hindus, the Mohammedans, the Parsis and the Christians, who have made India
their country, are fellow countrymen and they will have to live in unity, if only for
their own interest. In no part of the world are one nationality and one religion
synonymous terms: nor has it ever been so.’’28
Thus modern Indian political thinkers and leaders were forced to look for unity
of India in terms other than religious or linguistic unity. The starting point of such
a search for unity was a frank acknowledgment of the various forms of diversity.
Addressing the fifth meeting of the Indian National Congress held in Bombay in
1890, Ferozeshah Mehta expressed a fairly common sentiment at that time: ‘‘De-
spite social and religious differences, we have all begun earnestly to realize that we
are fairly on the way to a common national existence, united and bound together
by the common political ties.’’29
Insofar as the model of the nation-state was rooted in unease with diversity,
Indian nationalism could not but be deeply uneasy with this model. It could
not but reject images of essential unity drawn from a homogeneous cultural
and religious community. Writing just before Independence, Jawaharlal Nehru
summed up the nationalist thinking on Indian unity: ‘‘It was absurd, of course, to
think of India or any country as a kind of anthropomorphic entity. I did not do so.
I was also fully aware of the diversities and divisions of Indian life, of classes, castes,
religions, races, different degrees of cultural development. Yet I think that a
country with a long cultural background and a common outlook on life develops
a spirit that is peculiar to it and that is impressed on all its children, however much
they may differ among themselves.’’30
27. Ibid., p. 8.
28. M. K. Gandhi, Collected Works, vol. 10 (New Delhi: Publications Division, Government of
India, 1969), p. 23.
29. Extract of the speech in R. V. Ramachandrasekhara Rao, ed., Indian Unity: A Symposium
(New Delhi: Publication Division, Government of India, 1969), p. 21. This valuable official collec-
tion of extracts from the writings of a wide range of nationalist leaders reflects an anxiety on this
score that extended well into the postcolonial period.
30. Jawaharlal Nehru, Discovery of India (Delhi: Oxford University Press, 1981), p. 59.
54 crafting state-nations
Imagining a state-nation is necessary but not sufficient for crafting a viable state-
nation. This imagination needs to be translated into a set of robust political
institutions. Here again we encounter a specificity of the Indian experience: the
would-be crafters of the Indian constitution deliberated and reflected on the
institutional frame of independent India for nearly half a century. The search for
appropriate political arrangements for accommodation of diversities began long
before the Constituent Assembly met in 1946. In the very early days, the Indian
National Congress, institutional embodiment of the anticolonial struggle and the
dominant political party after independence, agreed on a consociational formula
for decision-making that gave Hindu and Muslim delegates a veto over every
major decision.31 Although the Congress Party rejected the colonial provision of a
separate electorate for the Muslims provided in the Government of India Act of
1909, it had already accepted the principle of special protection of minorities. By
the middle of the 1920s the Congress began detailed studies about the future
political set up of an independent India. The Congress Party rejected a British-
drafted Simon Commission Report and appointed its own committee under the
leadership of Motilal Nehru, the father of Jawaharlal Nehru, to suggest an outline
of a constitution for free India. The Nehru Report of 1928 foreshadowed many
31. See S. R. Bakshi, Dadabhai Naoroji: The Grand Old Man (Delhi: Anmol Publications,
1991), p. 105.
india as a state-nation 55
provisions of the Indian constitution. It shows that the critique of the colonial
definition of nationalism had already prepared the nationalist leadership in India
for some key elements of what we describe as the state-nation model. The exact
nature of these provisions were debated for the next two decades, but the core
elements of the state-nation structure were already articulated in the Nehru Re-
port, which was approved by the All Parties Conference in Lucknow in 1928.32
The definition of citizenship in the Nehru Report was very state-nation-
friendly in that it was absolutely inclusive and territorial: ‘‘the word ‘citizen’
wherever it occurs in this constitution means every person who was born, or
whose father was either born or naturalized, within the territorial limits of the
commonwealth’’ (article 3). It also laid down the propositions that the system of
independent India would be parliamentary (article 5) and bicameral and federal
(articles 8–9). We should note that a parliamentary federal system is the most
supportive combination for the emergence of ‘‘centric-regional’’ parties that may
be a useful alternative to ‘‘exit’’ for parties with many different linguistic major-
ities. The radical linguistic reorganization and reconfiguration of India that al-
lowed each Indian state to decide to govern itself in its local majority language
did not fully occur until 1957, but it was strongly supported at the 1921 Con-
gress meeting in Madras and endorsed in the Nehru Report: ‘‘The redistribution
of provinces should take place on a linguistic basis on the demand of the ma-
jority of the population of the area concerned’’ (article 86). This is a classic state-
nation ‘‘holding together’’ federal decision, unthinkable in a U.S.-style ‘‘coming
together’’ federation. Also supportive of a relatively strong ‘‘holding together’’
federalism was the provision for the Supreme Court of the Union (unlike the
‘‘coming together’’ federalism of the United States) to have ‘‘original jurisdiction’’
in almost all matters (article 49). On the all-important question of religion too, the
Nehru Report also supported state-nation policies. In the section on ‘‘Fundamen-
tal Rights,’’ it clearly ruled out an established religion, supported a religiously
impartial state but, unlike the U.S. Constitution, implied the admissibility of
some state aid for religious educational establishments (article 4).
Thus, when the Constituent Assembly formally set about its tasks in 1946, there
was little doubt about the provisions for protection of linguistic, cultural, and
religious diversity. The institutional structures the Constituent Assembly estab-
lished provided for a robust state-nation. In this sense, India’s constitution inher-
32. All Parties Conference (India). Nehru Committee, The Nehru Report: An Anti-Separatist
Manifesto/The Committee Appointed by the All Parties Conference, 1928 (New Delhi: Michiko and
Panjathan, 1975).
56 crafting state-nations
Yet the constitution was not sufficient to put India firmly on the path to build-
ing a state-nation. As Sunil Khilnani notes, the role of political leadership was
crucial: ‘‘The Indianness outlined in the two decades after 1947 was an extempo-
rised performance, trying to hold together divergent nationalism that resists sum-
mary in clear or simple doctrinal statements. It tried to accommodate within the
form of a new nation-state significant internal diversities; to resist bending to the
democratic pressures of religion; and to look outwards. This experimental re-
sponse to the question of how to be Indian was not a victory of theoretical consis-
tency. It was a contingent acquisition, based on a coherent if disputable picture of
India. It did not reassure itself by relying on a settled image of the culture, nor did
it impose one. That was its most important trait: it did not monopolize or simplify
the definition of Indianness.’’33
In the next two chapters we go into the details of how this constitutional design
and political approach was deployed in specific situations involving some of the
most serious challenges to the Indian state. Let us begin here by asking: Was this
33. Khilnani, Idea of India, p. 179. A recognition that Indian leadership was following an
unusual approach to ‘‘nation-building’’ was quite widespread among perceptive analysts of Indian
politics. Rajni Kothari argued that the Indian model of ‘‘nation building,’’ involving recognition and
political articulation of diversities, was relevant to many postcolonial societies. See his introduction
to Rajni Kothari, ed., State and Nation Building: A Third World Perspective (Bombay: Allied, 1976).
Rasheeduddin Khan argued that Indian federalism was unique, for the political institutions of
federalism were superimposed on a federal society. See his ‘‘Federalism in India: A Quest for New
Identity,’’ in Rethinking Indian Federalism, ed. Rasheeduddin Khan, ed. (Shimla: Indian Institute of
Advanced Study, 1997). Ravinder Kumar, a leading historian of modern India, contends that India
was a civilization-state rather than a nation-state. Kumar, ‘‘India: A ‘nation-state’ or ‘civilisation-
state’?’’ in Journal of South Asian Studies 25, no. 2 (2002), pp. 13–32.
india as a state-nation 57
‘‘idea of India’’ described above confined to the high traditions of political theory
and legal constitutional texts? Did this idea find resonance across different reli-
gions, regions, communities, and classes among ordinary Indian citizens? Or did it
remain an unrealized aspiration? To find out, we draw on empirical evidence in
the form of available surveys of political opinion, attitudes, and values.
One important indicator of identification of a citizen with the society and the
state is the sense of pride, in this case the pride in being Indian. Fortunately, a
comparable question has been asked in all the longstanding federal democracies
we are assessing and is continuously asked in the member states of the European
Union by the Eurobarometer surveys.
The pride question has been asked in the widely used World Values Survey
under the direction of Ronald Inglehart and his colleagues at the University of
Michigan. We have used here three of the five rounds of these surveys available
for our comparative analysis, 1990–93, 1995–97, and 1999–2001. The first round
covered 42 independent countries, the second 53, the third 75. India has been
included in all of these rounds. This set of surveys is particularly interesting for
comparativists, because respondents in each country are asked mostly the same
questions and many of the questions have been used in all of the rounds.34 We
have from the World Values Survey three fairly similar readings on the pride
question for India. In all three waves, about two-thirds of the respondents said they
were ‘‘very proud’’ to be an Indian. About 20% to 25% said they were ‘‘proud,’’ thus
pushing the proportion of ‘‘very proud’’ and ‘‘proud’’ to between 85% and 90%. In
each of these waves, the proportion of those who said they were ‘‘not proud’’ or
‘‘not at all proud’’ was recorded at less than 10%. This can be confirmed with a
report based on a major survey carried out by the Centre for the Study of Develop-
ing Societies that has a larger and more representative sample.35 Better represen-
tation of rural and uneducated voters in the CSDS survey has led to a slightly
higher proportion of ‘‘do not know’’ responses, but the basic pattern of response is
the same as reported by the World Values Survey. We can thus be confident that
Indians on the whole are highly proud of their national identity.
The World Values Survey allows us to place these figures in comparative
34. See Pippa Norris and Ronald Inglehart, Sacred and Secular: Religion and Politics Worldwide
(Cambridge: Cambridge University Press, 2004), appendix A, pp. 243–246.
35. The survey, called State of Democracy in South Asia (SDSA), was carried out in the five
countries of South Asia: India, Sri Lanka, Pakistan, Bangadesh, and Nepal. Yadav was the co-
principal investigator Stepan and Linz served on the advisory board and, working with Yadav,
created a number of new questions that were particularly relevant to the inquiry into state-nation,
such as multiple and complementary identities and religion and politics. For a summary of results
from the survey, see SDSA Team, State of Democracy in South Asia.
58 crafting state-nations
table 2.2
Pride in Nationality in the Eleven Longstanding Federal Democracies
Do not know/
Democracy Very proud Quite Not very Not at all not applicable
United States 71% 23% 4% 0% 2%
Australia 70 23 2 0 5
India 67 21 5 2 5
Canada 65 28 3 2 2
Argentina 65 24 4 3 5
Brazil 64 19 14 2 1
Spain 51 36 6 3 4
Austria 50 37 6 2 5
Belgium 20 46 15 7 11
Switzerland 23 47 16 7 7
Germany 15 46 22 7 9
Source: Data for all countries are from responses to the question ‘‘How proud are you to be
[nationality]?’’ Ronald Inglehart et al., eds., Human Beliefs and Values: A Cross-Cultural
Sourcebook Based on the 1999–2002 Values Survey (Mexico D.F.: Siglo XXI Editores, 2004).
Note: Percentages may not add up to 100 due to rounding.
perspective. Table 2.2 presents responses to the pride question across eleven long-
standing federal democracies from the latest wave of the World Values Survey. This
wave of World Values Survey (1999–2002, held in India in 2001) showed that only in
the United States (71%) and Australia (70%) was the proportion of those who feel
‘‘very proud’’ even marginally higher than in India (67%). The proportion of those
who felt ‘‘very proud’’ was lower in Argentina (65%), lower in Brazil (64%), lower
still in Spain (51%), much lower yet in Germany (15%), and surprisingly low in
Switzerland (23%).36 Those saying ‘‘not very proud’’ or ‘‘not at all proud’’ add up to
7% in India, certainly more than the United States (4%) and Australia (2%), but
lower than the other federal democracies, particularly Germany (29%), and sur-
prisingly again, Switzerland (23%). In the case of Germany, the Nazi period and
the Holocaust legacy represent a heavy burden, making it difficult for its citizens
to feel fully proud of their nation.37 Those who feel an exclusive identity with
another nation and reject the state in which they live are also not likely to feel
proud of the state or the nation and its heritage. Some of the variations across
36. On Switzerland, as the reader will see in table 2.9, consistent with our idea of state-nation, of
the eleven longstanding federal democracies, Switzerland has the highest percentage of people with
confidence in the government and the second-highest percentage of people with confidence in the
legal system. So while they may not have inordinate pride in being Swiss as such, they nevertheless
have great pride in their Swiss institutions.
37. On the complex issue of pride (or lack thereof ) in the German nation, see Elisabeth Noelle-
Neumann, ‘‘Nationalgefühl und Glück,’’ in Noelle-Neumann and Renate Köcher, Die verletzte
Nation (Stuttgart: Deutsche Verlags-Anstalt, 1987), 17–74.
india as a state-nation 59
table 2.3
Pride in India for All Citizens and for Marginal Groups, 2005
All of Scheduled Scheduled
India Muslim Caste Tribe Nonliterate
Very proud 60% 57% 56% 44% 44%
Proud 29 31 29 37 34
Not proud 2 1 3 2 3
Not at all proud 1 2 2 1 2
Do not know/no answer 8 9 10 15 17
(N) 5,227 636 901 427 1,964
Source: State of Democracy in South Asia: A Report (Delhi: Oxford University Press, 2008).
Note: Percentages may not add up to 100 due to rounding.
countries reflect these different attitudes. Thus, whether seen in its own terms,
over a period of time, or in comparison with other countries, Indian citizens’ sense
of pride in their nationality is quite impressive.
Aggregate figures are not, however, very instructive in dealing with question of
subjective identity. While it is important to understand the views of the citizens as
a whole, it is equally important to look at the views of any significant minorities
that may vary from this response. Therefore it is important to ask if the conclusion
about the high level of national pride holds across the different sections of society.
In order to find out, we need a disaggregated picture of national pride for some of
most salient marginal groups in India: Muslims (the largest and most disadvan-
taged religious minority), Scheduled Castes or Dalits (the ex-untouchable com-
munity that has historically suffered social exclusion), Scheduled Tribes (forest-
dwelling indigenous communities that have remained on the periphery of mod-
ern development), and nonliterates.
As table 2.3 makes clear, responses from the Muslim community and the
Scheduled Castes are no different than the average national response on this
question. The respondents from Scheduled Tribes and those without any formal
education do report lower levels of pride, partly because a larger proportion of
these groups fails to understand the question (itself a reflection of the uneven
dissemination of the idea of nationalism in modern India) and partly because they
are genuinely less enthusiastic about Indian nationalism. Yet this lower level of
enthusiasm does not lead to any significant rejection of the national identity: in
none of these groups does the proportion of those who do not feel proud at all
exceed 2%. Thus, notwithstanding the variations in the reception of and enthusi-
asm for the national identity, there is no noticeable rejection of the national
identity in any minority group for which we have information.
60 crafting state-nations
The sense of pride in nationality is a useful but limited vantage point to understand
the nature of cultural and political identities in any country. By posing the alterna-
tives in terms of pride or lack of pride, the question limits our ability to find out
much about the content of this pride or its absence. Citizens may attach very
different meanings to the India that they take pride in. For most Indian citizens,
India is a nation, so they see the Republic of India as a nation, though there are
divergent visions of that nationhood, some which would exclude from the nation
people that feel they are Indian without having to share an exclusivist conception of
the nation.38 There are many people who feel other identities, sometimes equally
strong, sometimes somewhat stronger than the Indian national identity. With the
data available to us, it is not easy to define how many people feel an Indian identity,
and even less easy to define those who have different conceptions of the Indian
nation. It is more difficult than in the case of Spain, since the matter of language, so
important in that case, is less hegemonic and defining in India. The linguistic
states recognize and allow for a certain identification with a distinctive language
and culture, and we quickly see that many Indians have a dual identity, that of
being Indian and that of belonging to their state. We discover that some of them feel
an identity with their state that may actually be a linguistic cultural identification,
without explicitly seeing themselves as Indian. It is not easy to say if that identity is
similar to that of a national identity in multinational societies.
While it is imperative to recognize areas of alienation in India, we also should
look at aggregate data on subjective political identities in India. Survey evidence
gathered by the CSDS allows us to examine the question of political identities in
contemporary India in an empirical manner. A question about subjective national
identity was asked in the Indian component of the State of Democracy in South
Asia Survey: 2005. Unlike both Spain and Belgium, the modal response category
in India is ‘‘only Indian.’’ Nearly half the respondents privilege their Indian iden-
tity over identities of their respective states, which in most cases also happen to be
linguistic identities.39 Even when asked about the in-between category of ‘‘more
38. On visions of the nation in India, see Ashutosh Varshney, ‘‘Contested Meanings: India’s
National Identity, Hindu Nationalism, and the Politics of Anxiety,’’ in Daedalus 122 (Summer 1993),
pp. 227–261.
39. While there is a great deal of overlap between the linguistic divisions of India and the
boundaries of the various states of the Indian union, it is not a perfect match. Most of the states have
significant linguistic minorities. The largest language, Hindi, is the official language of nine states of
the Indian union in addition to two Union Territories. Most of the large languages are internally
divided among various dialects that vary considerably.
india as a state-nation 61
table 2.4
Subjective National Identity and Religion in India
Hindu Muslim Christian Sikh All of India
Only Indian 34% 43% 30% 44% 35%
More Indian than state identity 12 11 16 9 12
As Indian as state identity 19 16 16 13 19
More state identity than Indian 10 7 18 17 10
Only state identity 12 10 13 16 12
Do not know/no answer 13 13 7 1 12
(N) (4,274) (636) (152) (87) (5,385)
Source: State of Democracy in South Asia Survey, 2005, CSDS.
Indian than state identity,’’ as many as 35% of all respondents chose ‘‘only Indian’’
to describe their identity, with 12% opting for the new category. To recall, only 13%
of Belgians and 14% of Spanish respondents opted for only national identity in
response to the same question. This reflects the depth of nationalist sentiment in a
country where nationalism emerged from intense anticolonial struggle. At the
same time, this response does not rule out significant reservations: about a fifth of
the respondents privileged their state identity over their Indian identity. Even with
the in-between option of ‘‘more state identity than Indian,’’ the proportion of those
who opted for ‘‘only state’’ identity was a little higher than in Belgium or Spain.
Less than one-fifth of Indians opted for the middle category of ‘‘equally Indian and
state identity,’’ compared to about half of the respondents in Spain and Belgium
who had equal and dual identities. Table 2.4 presents data from the 2005 State of
Democracy in South Asia (SDSA) survey for the whole of India and also for major
religious communities.
The implications about religion are quite clear: more Muslims—the largest
and the most disadvantaged religious minority in the country—identify them-
selves as ‘‘only Indian’’ than do the majority Hindu community. This is largely
because, except for the Kashmir Valley, India’s 161 million Muslims are excep-
tionally territorially dispersed and, again excepting Jammu and Kashmir, are not a
majority in any state of the union. Conceptually and politically, this means two
things. First, the great territorial dispersion of Muslims means that the vast major-
ity of Muslims cannot be accommodated by the mechanism of asymmetrical
federalism alone, meaning that other diversity-accommodating policies, such as
the equal support of all religions and some consociational devices, must contrib-
ute to engendering a state-nation. Second, Muslims self-identify more with their
Indian than their state identity.
62 crafting state-nations
The Christians, on the other hand, are more concentrated in certain pockets of
southern and northeastern India and constitute a majority in several small hill
states in the northeast (Meghalaya, Mizoram, and Nagaland). The Sikhs are split
into those who live in the Punjab and see themselves as primarily Punjabis and
those who live outside the Punjab and identify more with India. Here again, as in
analysis of pride by social groups, the central point is that notwithstanding differ-
ences among groups, there are no signs of serious disaffection with the political
community of India.
We would like to argue that democracy, and with it democratic political in-
stitutions and processes, are an essential component of the viability and stability
of multicultural, multilingual, multinational state-nations. Democracy makes
possible the identification with the state for many of its citizens who might
have different identities, who might question a nation-state, but are neverthe
less ready to be loyal citizens of the state. Authoritarianism might serve to im-
pose a nation-state model on the society, as was the case of Spain under the
authoritarian regime of Franco. But, as the data for Spain show, the result ul-
timately was a backlash of resurgent and, in the Basque Country, violent extreme
nationalism.
Democracy, and more concretely federal democracy, can serve to integrate
such a society. The Verfassungspatriotismus (‘‘constitutional patriotism’’) of Dolf
Sternberger, developed by Jürgen Habermas in the German Federal Republic, is
an important component for heterogeneous societies. It is not the only compo-
nent of support for a state-nation but certainly one of the most important ones. It is
for that reason that a more detailed discussion of the attitude of Indians and of the
various groups within Indian society toward democracy and democratic institu-
tions is vital to our interpretation of the politics of diversity.
In analyzing the attitudes of citizens toward the democratic institutions—be-
ginning with democracy, compared to authoritarian alternatives, and ending with
political parties—there are some serious problems with the indicators used. There
is sufficient evidence from many countries that we need to distinguish the belief
in the need for certain institutions and their desirability and legitimacy from the
attitudes about the way those institutions actually perform. People in principle
might support democracy, but they may often have serious misgivings about how
their particular democracy is actually functioning. The distinction between the
india as a state-nation 63
40. This is a theme developed in Juan J. Linz, ‘‘Crisis, Breakdown, and Reequilibration,’’ in The
Breakdown of Democratic Regimes, ed. Linz and Stepan (Baltimore: Johns Hopkins University Press,
1978).
41. For an elaboration on this theme, see Juan J. Linz, ‘‘Parties in Contemporary Democracies:
Problems and Paradoxes,’’ in Political Parties: Old Concepts and New Challenges, ed. Richard
Gunther, José Ramón Montero, and Juan J. Linz (New York: Oxford University Press, 2002), pp.
291–317.
64 crafting state-nations
Brazil, and Korea but clearly behind Spain and Uruguay. Another interesting fact
from table 2.5 is that the proportion of Indians who support authoritarianism (6%
and 4% in 1999 and 2004, respectively) or are indifferent to this choice (7% and
6%) is very small, even when compared to Spain and Uruguay. The unusually
large proportion of the ‘‘do not know’’ response (27% and 20%), the highest
recorded in any country for this question, accounts for the rest.
Given the strong relationship between a low level of education and saying ‘‘do
not know’’ to this question, it is reasonable to assume that the ‘‘do not know’’
response is more a function of an inability to comprehend words like democracy
and authoritarianism rather than unarticulated reservations about democratic
political systems.42
Given this strong ‘‘do not know’’ effect and its correlation with education, a
robust comparison of the Indian data with other countries would require neu-
tralizing this effect. The figures in parentheses in the row for support for democ-
racy achieve this by looking at the percentage of the support for democracy
among those who give a valid response; the ‘‘do not know’’ response is treated as
missing data in this calculation, a common practice in statistical analysis. Thus
the level of support for democracy was 83% and 88%, respectively, in the 1999 and
2004 surveys, about the same as Uruguay and Spain, if we take only valid re-
sponses into account. In sharp contrast, Brazil, Chile, and Korea, among valid
responses, range between 41% and 58%.
In Brazil, those clearly preferring democracy are only 41%, with a substantial
proportion expressing potential support for an authoritarian alternative or no
preference, a much less favorable response than we find in India. The figures for
Spain, a relatively new democracy, are very similar to those we find in India. The
Spanish data have been consistently within this range over a long period of time
and are very similar to those we find in the other southern European democracies.
Of the twenty countries of Latin America, only Uruguay is regularly more support-
ive of democracy than India, although Costa Rica sometimes places high as well.
Even in Chile, after the painful experience of authoritarianism, democracy is
endorsed by only 52%, with a large proportion of those who are indifferent. In
Korea, 27% of respondents, in contrast to 4% in India, agreed with the statement,
‘‘In some circumstances an authoritarian government can be preferable to a
42. In 2004 the overall proportion of the respondents who said ‘‘do not know’’ to this question in
India was 20%. The proportion was 4% among those who had completed graduation or a higher
degree, 9% among those who had completed high school, 19% among those who managed to
complete primary education, and 37% among those who had no formal education.
india as a state-nation 65
table 2.5
Attitudes toward Democracy and Authoritarianism in India and
Five Important ‘‘Third Wave’’ Democracies
India India
Attitude Uruguay Spain (2004) (1999) Korea Chile Brazil
Sources: Data for India are from National Election Study, 1999 and 2004, Center for the Study of Developing
Societies, Delhi. Data for Uruguay, Brazil, and Chile are from Latino Barometer 1996, directed by Marta
Lagos. Data from Spain are from Eurobarometer 37 (1992). Data from Korea are from Korea Democracy
Barometer, 2004, directed by Doh Chull Shin.
43. The much greater openness to nondemocratic solutions expressed by respondents in Korea,
Chile, and Brazil in contrast to India needs more research and reflection. However, a hypothesis
worth particular study is previous experience with and perception of nondemocratic and demo-
cratic regimes. For a significant section of the Chilean voters, the center right, the Pinochet military
experience is associated with providing greater economic growth and more personal security. In
Korea, the military authoritarian rule is associated with high economic growth. In Brazil, for many
poor people, the most popular president was probably Getúlio Vargas, whose periods of rule were
often marked by a more ‘‘inclusionary corporatist’’ style of populist authoritarianism than democ-
racy per se. Independent India, in contrast, has never had a military regime, so it is not seen as an
acceptable or desired alternative. The low support for nondemocratic alternatives in Uruguay may
also be partly because the military is not associated with any successful and popular period of rule
and they had no ‘‘inclusionary corporatist’’ figure comparable to Vargas.
66 crafting state-nations
table 2.6
Support for Democracy in India by Four Disadvantaged Social Groups, 2004
All of Scheduled Scheduled Very
Attitude India Caste Tribe Poor poor
weaker among the marginalized sections of society. Table 2.6 offers this necessary
disaggregation for four major disadvantaged social groups (the ex-untouchables of
the Hindu social order or the ‘‘Scheduled Castes,’’ the indigenous people or the
‘‘Scheduled Tribes’’) and the two poorest socioeconomic groups.
The most striking thing about the data in this table is that there is not a single
exception to the national norm of very high support for a democratic political
system. In no group does the support for democracy fall below the 60% mark, and
nowhere does the support for authoritarianism reach double digits. At first sight,
some groups appear to deviate slightly from the national norm: the support for
democracy among the Scheduled Caste, the Scheduled Tribes, and the Very Poor
is about five to ten percentage points lower than the national average. But much of
this is a function of larger proportion of ‘‘do not know’’ among these groups, which
is directly associated with lower levels of educational attainment. Once we control
for ‘‘do not know’’ responses—by examining the valid percentages after treating
‘‘do not know’’ as missing data—most of the apparent, even if minor, differences
among various groups and states disappear. On average, 88% of those respondents
who gave valid responses support democracy; all but one group (the Scheduled
Tribes, which register 81% valid responses for democracy) fall within five percent-
age points of this national average.
the ‘‘clash of civilizations’’ literature would suggest? India has traditionally been a
country with high levels of religious heterogeneity, coupled with high levels of
religious practice and belief. According to the 2001 census, India had approxi-
mately 805 million Hindus, 138 million Muslims, 2.3 million Christians (about
60% Catholic and 40% Protestant), 800,000 Buddhists, 400,000 Jains, 600,000
‘‘other religions,’’ and only 100,000 people who did not state a religion.44 India also
has some of the highest levels of religious belief and practice in the world; 93% of
the population describe themselves as believing in God, and 87% as being ‘‘very’’
or ‘‘somewhat’’ religious, 53% as praying daily; almost half (at least 400 million
people) say that they have gone on a pilgrimage or traveled to another place for
religious purposes within the last ten years. Finally, against this very high base,
nearly four times as many respondents say that in the last ten years their ‘‘family’s
engagement in religious activities’’ has increased as say that they have decreased.45
In recent decades, contrary to classic modernization theories, India has combined
steadily increasing levels of education and urban living with sharply increasing
self-reported levels of religious practice, especially among the more urbanized
and better educated. For example, 46% of villagers say they pray daily, whereas
65% of those in towns and 77% of city-dwellers say they pray daily. The same trend
is visible in education: 44% of nonliterates say they pray daily, but 56% of univer-
sity graduates say they do, and whereas 18% of nonliterates were classified as ‘‘high
participants’’ in public religious activities, 24% of university graduates were so
classified.46
In such a context, what choices concerning the state’s policies toward religion
might have seemed most, and least, attractive to members of the India Constitu-
ent Assembly if they were interested in social peace, some social reform, and
enlisting the support of the vast majority of this heterogeneous population for a
democracy?
A small group of Hindu nationalists wanted to impose Hinduism as the estab-
lished religion in India. An established church is not necessarily undemocratic.
Arguably, for the last sixty years of the twentieth century, the countries in the
44. Census of India, 2001. Since these data are already old, we have in this book occasionally
used more recent estimates from other sources, especially for the population of Muslims.
45. The above percentages concerning religious beliefs and practices are from the State of the
Nation Survey, carried out by the CSDS in January 2007. The survey used a national representative
sample of 15,373 respondents. Figures reported here are for questions B5, B3, B11, B6 and B17,
respectively, and are sourced from the Data Unit, CSDS, Delhi.
46. Ibid. For a rich analysis of the data in this report, see Sanjay Kumar, ‘‘Religious Practices
among Hindus: Does This Influence Their Political Choices?’’ CSDS Working Paper, Delhi, June
2008, pp. 3–9.
68 crafting state-nations
world with the highest quality of democracy, welfare, equality, and political sta-
bility were the Scandinavian countries. But we must remember that every single
one of these high quality democracies had an established religion—Evangelical
Lutheranism—during this entire period.
What conditions helped make such establishment socially and religiously ac-
ceptable? Such establishment was almost certainly facilitated by Scandinavia’s
combination of high religious homogeneity (in Denmark, for example, as late as
1990 more than 96% of the population were Lutherans) with a low intensity of
religious practice (in Denmark, for more than forty years only about 5% of the
population reported going to church once a week). In contrast, none of the West
European countries that are substantially more religiously heterogeneous than
Scandinavia, such as modern Germany, Belgium, Holland, or Switzerland, has
an established religion.47
In sharp contrast to Scandinavia, precisely the opposite conditions existed in
India’s polity, which features low religious homogeneity and high religious prac-
tice. Less propitious still, in the post-Independence conditions of Partition be-
tween India and Pakistan, where even conservative estimates indicate that a half a
million people were killed, Muslims would have felt very threatened if India had
established Hinduism.
The French model of secularism, especially during the Third Republic (1870–
1940) in its 1905 laïcité form, was driven by an animus against the Catholic
Church, which many French secularists saw as a part of the ancien regime, and
indeed by their religiously hostile desire to make the French state ‘‘free from
religion.’’48 For many Indians—this included Gandhi and his followers—a re-
ligiously hostile secularism would not have helped generate deep and sustained
support for the new democracy. The U.S. version of the wall of separation be-
tween church and state would have made it difficult for the new democracy to
give financial aid to impoverished religious facilities such as those once supported
by the heads of the 565 princely states, which lost their political and financial
autonomy once they ‘‘acceded’’ to the union after Independence and which had
47. See Stepan, ‘‘Multiple Secularisms.’’ For a pioneering discussion of Europe’s ‘‘historic
mono-confessional belts’’ versus ‘‘multi-confessional belts’’ and the political consequences of these
differences, see John T. S. Madeley, ‘‘A Framework for the Comparative Analysis of Church-State
Relations in Europe,’’ West European Politics 26, no. 1 (January 2003), pp. 23–50.
48. See Ahmet Kuru, Secularism and State Policies toward Religions: The United States, France,
and Turkey (Cambridge: Cambridge University Press, 2009), esp. chs. 1, 4, and 5. Also see his
‘‘Secularism, State Policies, and Muslims in Europe: Analyzing French Exceptionalism,’’ Compara-
tive Politics 4 (2008).
india as a state-nation 69
been a part of Indian tradition for thousands of years. In any case, many Indian
leaders were not attracted by an aspect of the U.S. model, which is the opposite of
the French model in that it appears to be driven by a desire to attain ‘‘freedom of
religion from the state.’’49 For example, the chair of the drafting committee of the
constitution for the Constituent Assembly, the ex-untouchable B. R Ambedkar,
along with many others, felt that if some religious bodies were violating the rights
of the citizens, say by forbidding the ex-untouchables from entering temples, then
the new democracy should not have a ‘‘wall of separation from religion’’ but must
be allowed to help expand citizens’ liberties and dignity in cases of egregious
denial by religious authorities. The major theorist of Indian secularism, Rajeev
Bhargava, calls the Indian model of action vis-à-vis religions ‘‘principled distance’’
in that it is a ‘‘multi-value secularism’’ that respects all religions but does not rule
out the use of the democratic state to protect its citizens against some undemo-
cratic religious practices being imposed on them.50
In this historical and normative context, the model of state religion relations
crafted by the Indian Constituent Assembly and developed later was a highly
original model with strong affinities to our state-nation model. All religious com-
munities were recognized and respected by the state. All religious communities
could, for example, run schools, organizations, and charities eligible for state
financial support. The norms and practices of this model are so pervasively ac-
cepted that the Hindu nationalist Bharatiya Janata Party (BJP) did not dare, when
it was the head of the ruling coalition, not to honor the tradition of giving exten-
sive state subsidies to help Muslim citizens make the hajj to Mecca.
We do not want to make the case that the Indian model of secularism, by itself,
created the attitudinal and behavioral patterns we are about to present, but we do
believe that the state-nation-like policies and values of India’s secular model
helped Indians address their great religious heterogeneity and their great intensity
of religious practice and helped bring about the remarkable consensus among all
religions concerning democracy that we will document.
The first point we would like to stress is that the percentage of members of all
four of the major religions in India who self-identify as having a ‘‘preference for
democracy as opposed to any other system’’ is very high by world standards;
Muslims (87%), Hindus (88%), Sikhs, (88%) and Christians (91%). The Muslims,
table 2.7
Support for Democracy in India as a Whole and by Four Major Religions
Attitude All of India Hindu Muslim Christian Sikh
the largest and the most disadvantaged religious community in India, thus have
about the same figures as the national norm for all four of the categories concern-
ing democracy (see table 2.7).
Given the self-reported increase in religious practice in India noted previously,
we constructed an ‘‘intensity of religious practice index’’ ranging from ‘‘low’’ to
‘‘high’’ to see if the trend toward growing intensity of religious practice correlated
with growing undemocratic attitudes and practices, as some had feared. Based on
the data in figure 2.1, the exact opposite trend has occurred. For all four major
religions in India, a steady counterintuitive trend is discernable; for each increase
in the intensity of religious practice, there is an increase in support for democracy.
In the State of Democracy in South Asia survey, we constructed a battery of
questions exploring the relationship between religion and democracy. Unfortu-
nately, our sample size (5,387, compared with 27,189 for the National Election
Study, 2004) only permits us to do a detailed comparative study of the two largest
religions in India, Hinduism and Islam. The sample size for other minority reli-
gions is too small for a robust analysis. A key question we wanted to explore was the
relationship of increased levels of ‘‘intensity of religious practice,’’ our indepen-
dent variable, to four critical components of democratic political society, which
will be our dependent variables: ‘‘political efficacy,’’ ‘‘overall trust in political
institutions,’’ ‘‘satisfaction with the way democracy works in this country,’’ and
‘‘voting ratios.’’ As table 2.8 makes clear, again counterintuitively when set against
much of the literature, on all eight observations (Hindus and Muslims on each of
the four variables) the group with ‘‘high religiosity’’ has higher scores on each of
the four variables than does those the group with ‘‘low religiosity.’’
95
93
92
Support for Democracy
91
90
89
89 89
88 88
86 86
85
83 83
80
Low Medium High
Intensity of Religious Practice Index
Figure 2.1. In All Four of India’s Major Religions, the Greater the Intensity of Religious
Practice, the Greater the Support for Democracy
Notes: The analysis is based on valid answers in the National Election Study—2004
[India] (total n = 27,189). Valid responses for the table are: Hindus = 17,261; Muslims =
2,549; Sikhs = 544; and Christians = 697. The findings for Hindus are statistically
significant (Pearson’s chi square [0.001), which means that the probability of the findings
occurring by chance are less than 1 per 1,000. The findings for Muslims are also statistically
significant (Pearson’s chi square [0.050), which means that the probability of the findings
occurring by chance are less than 1 in 20. The findings for Sikhs and Christians are also
positive but not statistically significant. Valid answers exclude ‘‘Do Not Know’’ and ‘‘No
Opinion’’ from the analysis for reasons we have already explained. ‘‘Support for
democracy’’ (Q23) is as measured in table 2.7. The ‘‘intensity of religious practice index’’
was computed by adding the self-reported frequency of praying (Q34a), visiting a religious
place (Q34b), participating in religious meetings (Q35a), making donations to religious
organizations (Q35b), and fasting (Q35c). In the index, 50% of the weight is given to
frequency of praying, as it is the highest in all religions. Participating in religious meetings
and fasting are given 20% weight each, as both these activities are high among all the four
major religious groups in India but are much lower compared to the frequency of praying.
Going to a place of worship and making donations have been given 5% weight each
because making donations depends on the economic class of the respondent and because
many Muslim women do not go to mosques. To further analyze the effect of the intensity
of religious practice on support for democracy, we made a binary logistic regression model.
In addition to the intensity of religious practice, we added as control variables efficacy of
vote (q21), membership in organizations other than caste or religious organization (Q19),
whether the respondent voted in the 2004 parliamentary election (Q3), gender,
respondent’s education (B4), monthly household income (B19), and level of urbanity
(B10). The coefficient on the index of religiosity is 0.138. Thus, we can say that a one-unit
increase in the index of religiosity (controlling for other factors) predicts approximately a
3.5% increase in the probability of support for democracy.
72 crafting state-nations
table 2.8
Relationship between Intensity of Religious Practice and Support for Political Institutions in India
Support (%)
Religious Low Medium High Net gain from
group intensity intensity intensity low to high
Sources: The analysis is based on the State of Democracy in South Asia survey conducted in 2005 for ‘‘trust’’
(Question C-13 battery), voting (Question C-8), and ‘‘satisfaction’’ (Question C-12). ‘‘Efficacy’’ (Question 21)
is based on National Election Study [India], 2004.
Note: The findings for ‘‘efficacy’’ and ‘‘trust,’’ among both Hindus and Muslims, are statistically significant
using a Pearson’s chi-square test (p-value [ .001), which means that the possibility of the findings occurring
by chance are less than 1 in 1,000. ‘‘Trust in public institutions’’ is an index created by adding the responses to
the frequency of self-reported trust in ‘‘Central government’’ (C-13a), ‘‘Provincial government’’ (C-13b),
‘‘Local government’’ (C-13c), ‘‘Civil service’’ (C-13d), ‘‘Police’’ (C-13e), ‘‘Army’’ (C-13f ), ‘‘Courts’’ (C-13g),
‘‘Parliament’’ (C-13h), ‘‘Political parties’’ (C-13i), ‘‘Election Commission’’ (C-13j). The findings for ‘‘voting’’
are statistically significant for Hindus (p-value [ .001) and for Muslims (p-value [ .050). The findings for
‘‘satisfaction’’ for Hindus are statistically significant (p-value [ .001), while the findings for Muslims are
positive but not statistically significant.
The findings indicate that the followers of all four major religions in the state-
nation of India do not see any contradiction between their practice of religion and
their practice of democracy. It is fair to say that both democracy and religion are
integral and valued aspects of Indian citizens’ public and private lives.
trust in institutions
Although we will turn later to the problem of the distrust in some institutions
(particularly in political parties and the police), it is important to note how much
the average Indian agrees with the need for some of the key political institutions
india as a state-nation 73
necessary for a democracy. The CSDS has been using a somewhat different
question to tap the legitimacy of the Indian democratic polity: ‘‘Suppose there
were no parties or assemblies and elections were not held, do you think that the
government in this country could be run better?’’ The percentage of valid respon-
dents saying ‘‘no’’ increased from 74% in 1971 to 91% in 2004. This question is
valuable for double-checking the conclusions drawn above: it has the advantage
of measuring popular attitudes to democracy (without using the ‘‘D-word’’ itself )
and also of avoiding a lazy ‘‘yes’’ by requiring the supporters of democracy to
disagree with the statement.
Since political parties are one of the key democratic institutions, it is also
useful to examine the level of party identification in India through the CSDS
surveys over the last four decades (1971, 1996, and 2004). A similar movement
across time can be observed in levels of party identification. Notwithstanding
widespread popular disenchantment with political parties, the proportion of those
who felt ‘‘close to a political party’’ in 2004 was quite high by global standards and
up from the level recorded in 1971. While the global trend is in the direction of
disengagement from political parties, the Indian data suggest otherwise.
Let us try to go beyond the overall question of support for the principle of
democracy to trust in the functioning of the existing institutions of the Indian
state. The study of trust has been a major research area in policy analysis and
social science research for the last twenty years or so. Pippa Norris attempted to
bring this research together so as to be able to make comparative judgments about
trust in institutions.51 Using the World Values Survey 1990–93 round, she con-
structed a composite index of trust by measuring expressed trust in six major
institutions she felt were important for a democratic state: parliament, the legal
system, civil service, the judiciary, police, and the military. She did this for twenty-
one democracies, one of which was India. India’s composite score for citizen’s
institutional trust was the highest among the twenty-one democracies, decisively
ahead of its closest competitor, Norway (see figure 2.2).
One has to exercise caution in interpreting a composite index based on mea-
surement at one point in time. In order to see if India’s very high comparative
standing concerning trust in institutions held up in a later survey, we checked the
data on trust in institutions for the eleven longstanding federal democracies based
on the 1995–97 round of the World Values Survey. If we combine the total
percentages of respondents who answered that they had a ‘‘great deal of trust’’ or
51. Pippa Norris, Critical Citizens: Global Support for Democratic Governance (Oxford: Oxford
University Press, 1999), esp. p. 229.
74 crafting state-nations
14
13
12
11
10
ic
No dia
Ire ay
Br d
nm n
k
A
Ca ile
Sw ada
Fin en
Ne r m d
er y
ng s
Fr ry
Au ce
Ja ia
S n
Po pain
e c e l al
Re m
ly
Hu nd
th a n
ar
D e tai
pa
n
G e lan
US
bl
Ita
r
Cz B tug
a
h giu
rw
an
Ch
st
ed
la
In
pu
la
n
i
r
Figure 2.2. Institutions and Political Trust in India and Twenty Other Democracies,
1990–1993
Source: Pippa Norris, Critical Citizens: Global Support for Democratic Governance
(Oxford: Oxford University Press, 1999), figure 11.2, p. 229.
‘‘quite a lot of trust’’ in an institution, India ranks first or second out of the eleven
longstanding federal democracies in five of the six categories. The only exception
was, unsurprisingly, the trust in the police, where India ranks second-last, above
only Argentina. For a poor and diverse country in which over a third of the citizens
are not literate and over a quarter live below the poverty line, such a low level of
trust in police is very disturbing. It suggests an unequal access to the rule of law
that compromises the quality of democratic governance the citizens can enjoy.
This lack of trust in police becomes critical for minorities and other marginalized
groups in situations of group violence; the Indian police have often been accused,
not without substance, of inaction and complicity during times of violence against
minorities and marginal groups.52 Despite this low trust in the Indian police,
Indian citizens, in contrast to Pakistani citizens, to our surprise, still see the police
as part of a ‘‘useable state.’’ In the State of Democracy in South Asia survey, in
answer to the question: ‘‘If you had a problem that needed police help, would you
52. Some of these instances are the massacre of Sikhs in Delhi and rest of the country following
the assassination of Indira Gandhi in 1984, the large-scale killings of Muslims in Meerut, Uttar
Pradesh, in 1987 and the massacre of Muslims in Gujarat in 2002. In all these instances, the police
force was either a silent spectator to mass killings or was complicit in the crime. These and other
lapses on the part of the police are extensively documented by human rights organizations and
groups such as those found in footnote 21 in this chapter.
india as a state-nation 75
go to the police?’’ 82% of Indian respondents answered ‘‘yes,’’ whereas only 27% of
Pakistani respondents answered ‘‘yes.’’53 From the perspective of trust, or lack of
trust, in the state security apparatus, we should note that the lack of trust in the
regular police force is not reflected in the case of the armed forces. The Indian
Army, which exceptionally in the developing world has never deposed a govern-
ment in a coup or ruled the country, is one of the most trusted institutions in
India.54
With the exception of the police, the next institution the Indians trust least is
political parties. But, this abstract disaffection does not prevent them from identi-
fying with specific political parties in greater numbers than elsewhere. Besides,
Indians trust political parties more than citizens in any other longstanding federal
democracy do (see table 2.9).
We can conclude our discussion on trust here except for one technical aspect.
The conclusions so far are drawn from the World Values Survey data. The Indian
sample within the World Values Survey underrepresented nonliterates and rural
dwellers in the rounds we have used and was only carried out in eight of the
country’s then fifteen official languages. Therefore we need to double-check these
conclusions with the help of a national representative sample such as those car-
ried out by the CSDS to measure trust with greater confidence. It is important to
check if the inclusion in the sample of more nonliterates and more rural dwellers
presents a somewhat different picture than that presented by the World Values
Survey. Table 2.10 presents a comparison of the responses on trust in institutions
from the World Values Survey of 2001 with the SDSA survey carried out by the
CSDS in 2005. While the level of trust recorded in the latter is a little higher than
that in the World Values Survey, the internal order of the five institutions is more
or less the same: police and political parties finish at the bottom of the heap in
both surveys, and the central government is at the top of both surveys. The data
from the survey carried out by the CSDS have the advantage of expanding the
range of institutions about which we have information. It shows that two of
the most trusted institutions in India are the judiciary and the Election Com-
53. State of Democracy in South Asia survey, Question 38, data sourced from Data Unit,
CSDS.
54. For example, in Pakistan only 38% expressed ‘‘great trust’’ in the military, in Sri Lanka only
31% did so, whereas in India 55% affirmed ‘‘great trust.’’ See SDSA Team, State of Democracy in
South Asia, pp. 250–251. One should also note that India is exceptional among developing countries.
Since World War II, 18 of the 20 countries in Latin America have experienced direct military rule. In
sub-Saharan Africa, more than 30 newly independent countries have had major military involve-
ment in their governments. In South and Southeast Asia, Indonesia, Korea, the Philippines, Paki-
stan, Thailand, Bangladesh, and arguably Taiwan have had long periods of military rule. The
military has governed Pakistan for about half the years since Independence.
table 2.9
Citizen Trust in Six Major Institutions in Eleven Longstanding Federal Democracies
India Switzer- United Bel- Ger- Aus- Argen-
Institution (rank) land Canada Brazil Austria States gium Spain many tralia tina
Legal system 67% (1st) 65% 54% 55% 58% 36% 45% 45% 54% 35% 27%
Parliament 53 (1st) 41 38 33 41 30 43 35 28 31 16
Political parties 39 (1st) 25 n.a. 32 n.a. 21 n.a. 18 14 16 8
Central government 48 (2nd) 50 38 48 n.a. 31 n.a. 30 24 26 27
Civil service 53 (2nd) 43 50 59 42 52 43 40 47 38 8
Police 36 (10th) 67 84 45 68 71 51 61 70 76 23
Sources: Data for all countries but Austria, Belgium, and Canada are from Ronald Inglehart et al., World Values Survey: 1995–97, Inter-University Consortium
for Political and Social Research, University of Michigan. Data for Germany are from the Länder of the former West Germany. Canada, Belgium, and Austria
were not included in the 1995–97 survey. Data for these countries are from World Values Survey: 1990–93. For both the 1990–93 and 1995–97 surveys, the
question numbers were 137 and 141 to 145. Question 143 was not asked in Canada. Questions 142 and 143 were not asked in Belgium and Austria.
Note: All figures in percentages, rounded to the nearest integer. Table entries refer to the sum of those who respond ‘‘A great deal’’ or ‘‘Quite’’ to the question
‘‘How much confidence you have in . . . ? (Great Deal, Quite, Not Very Much, None).’’ The countries are presented in the above table from left to right in
descending order of average trust. Thus, India has the most trust, on average, and Argentina the least.
india as a state-nation 77
table 2.10
Trust in Institutions in India
State of Democracy in South Asia, World Values Survey,
Institution 2005 2001
Central government 62% 56%
Legal system 59 n.a.
Election commission 51 n.a.
Civil service 47 49
Parliament 43 55
Police 42 38
Political parties 36 34
(N) (5,385) (2,002)
Note: Table entries show percentage who responded that they had ‘‘a great deal’’ or ‘‘quite a lot’’
of trust in each of the institutions. The ‘‘N’’ value for the World Values Survey refers to the total
number of cases, but the percentages are calculated from valid cases.
55. The 2002 riots in Gujarat and the 2002 electoral outcome in Gujarat raised many disturbing
questions. However, the Election Commission emerged stronger in one respect: against the ruling
BJP desires, the Election Commission managed to postpone the timing of the Gujarat elections.
The Election Commission also gained prestige due to the positive role it played in conducting
reasonably fair elections in Kashmir in 2002.
78 crafting state-nations
69.7%. In four consecutive observations between 1994 and 2000, the average
percentage of respondents who felt efficacious was only 39.7%.56
The National Election Study series in India uses a slightly different question
for measuring political efficacy: ‘‘Do you think your vote has an effect on how
things are run in this country or do you think your vote has no effect?’’ Unlike the
data from the United States, we do not have a large number of data points, but the
Indian NES offers three observation points—1971, 1996, and 2004—that allow us to
discern a trend. The trend in India is quite the opposite of the United States: 48%
of the respondents felt efficacious in India in 1971, 59% felt efficacious in 1996,
and 68% felt efficacious in 2004. To be sure, efficacy is being measured here with
reference to the value of the vote, and the significant changes in this respect partly
reflect the growing competitiveness in India’s electoral politics in the last fifteen
years. Yet if a significantly greater proportion of citizens feels that their vote makes
a difference, that is an undeniable positive for democracy and resonates well with
the data on trust that we have examined.
Table 2.11 presents the data on changing levels of citizens’ sense of efficacy and
legitimacy of the Indian democratic system for some of the marginalized groups
from 1971 to 2004. In 1971 the eight groups that expressed the lowest efficacy were
Scheduled Tribes (ST), nonliterates, women, the lowest economic quintile (the
‘‘very poor’’), Scheduled Castes (SC), the second-lowest economic quintile (the
‘‘poor’’), the rural population, and Muslims. Since that time, each of these eight
groups has reported a substantial increase in its level of efficacy.
This growing sense of efficacy could have expressed itself in a growing desire to
be anti-system and with an increasing dissatisfaction with democracy. In India,
however, what we see between 1971 and 2004, is a high correlation between the
previously most marginal groups feeling more efficacious and expressing greater
support for democracy. From 1971 to 2004, every single one of the eight groups that
had scored the lowest on efficacy in 1971 increased its support for democracy by 27
to 35 points (see table 2.11). Once we control for the ‘‘do not know’’ effect, there are
no noteworthy social differences in support for the democratic system.
Do these positive attitudes on trust, efficacy, and legitimacy translate in terms
of behavior as well? There is a general presumption, especially in the U.S. socio-
logical literature, that the lower the education level and the lower the income
56. See Virginia Sapiro, Steven J. Rosenstone, and National Election Study, Center for Political
Studies, University of Michigan, The American National Election Studies Cumulative Data File,
1948–2004. The wording of the two key questions was: ‘‘People like me don’t have any say about what
the government does’’ (V613) and ‘‘I don’t think public officials care much what people like me
think’’ (V609).
india as a state-nation 79
table 2.11
Sense of Political Efficacy and Legitimacy of Democracy among
Marginalized Groups in India, 1971–2004
Political efficacy Support for democracy
Group 1971 1996 2004 1971 1996 2004
National average 48% 59% 68% 43% 69% 72%
Scheduled Tribe 31 48 59 41 66 68
Women 36 51 61 32 64 67
Nonliterates 36 47 55 31 62 61
Very poor 38 51 60 32 64 66
Scheduled Caste 42 60 65 38 67 69
Poor 43 55 68 37 68 71
Rural 44 57 66 39 69 70
Muslims 50 60 66 40 72 73
Source: National Election Study [India], 1971, 1996, and 2004, CSDS, Delhi.
Note: The efficacy question was ‘‘Do you think your vote has effect on how things are run in
this country or do you think your vote makes no difference?’’ The support for democracy
question was ‘‘Do you think that the government in this country can be run better if there are no
parties or assemblies or elections?’’
level, the lower the voting participation rates and the lower the sense of personal
political efficacy.57 This is coupled with the belief that in the modern world,
political trust in institutions has been declining for over three decades. Finally,
there is Samuel Huntington’s famous axiom that if participation increases faster
than institutionalization, there can be a crisis of governability. Most of these
assumptions and worries are true for the United States. Some are true for Western
Europe. None is true for India.
In the United States this assertion (the lower the socioeconomic status, the
lower the voting turnout) holds true with brutal regularity. Table 2.12 presents
some of this evidence for the United States and India. If we divide levels of U.S.
income into five quintiles, for each quintile that income decreases, there is a
monotonic decline in voter turnout: 77%, 67%, 59%, 52%, 43%. The same holds
true, even more sharply, for the six levels of education; the percentage of post-
graduate voters is 84%, then for each descending level of education the percent-
ages are 79%, 66%, 57%, 43%, 38%. Also, blacks and Latinos vote less than whites.
In the nonpresidential year of 1994, 47% of whites voted, 37% of blacks voted, and
20% of Latinos voted.
In sharp contrast to the United States, voting rates in India do not decline as
57. See, for example, Sidney Verba and Norman H. Nie, Participation in America: Political
Democracy and Social Equality (New York: Harper and Row, 1972), p. 97.
80 crafting state-nations
table 2.12
Socioeconomic Status and Voting Turnout in the United States and India
United States (1988) India (1998)
(1998 turnout, 49%; 1988 turnout, 60%) (total turnout, 62%)
Income Income
1) Lowest quintile 43% a 1) Lowest quintile 57.1% b
2) 52 2) 65.2
3) 59 3) 73.3
4) 67 4) 59.6
5) Highest quintile 77 5) Highest quintile 46.6
Education Education
No high school 38 Illiterate 56.5
Some high school 43 Up to middle school 82.6
High school graduate 57 College 56.5
Some college 66 Postgraduate 41.0
College graduate 79
Postgraduate 84
% Voted
1994 1996
Community (midterm year) (presidential year) Community 1971 1998
White 47 56 Hindu (upper) 61.4 60.2
Black 37 50 Hindu (OBC) 45.3 58.4
Latino 20 27 Scheduled Caste 57.5 75.1
Scheduled Tribe 35.9 59.0
Muslim 87.9 69.6
Sikh 84.6 89.4
Sources: For India, Yogendra Yadav, ‘‘Electoral Politics in Time of Change: India’s Third
Electoral System, 1989–99,’’ Economic and Political Weekly, vol. 34, August 21–28, 1999, pp.
2393–99. For the United States, Jan E. Leighley and Jonathan Nagler, ‘‘Socioeconomic Bias in
Turnout, 1964–1988: The Voters Remain the Same,’’ American Political Science Review 86
(September 1992), pp. 725–736.
a
U.S. turnout expressed as percentage of voting-age citizens.
b
India’s turnout expressed as percentage of the national electorate.
one goes down the social hierarchy. In the 1998 Lok Sabha election, for example,
nonliterates had a higher turnout than did postgraduates, who in fact reported the
lowest turnout rate. In terms of income, the two poorest quintiles voted at substan-
tially higher rates than did the wealthiest quintile. The same pattern has been
repeated in virtually every election in the recent past. In terms of ethnic or
cultural communities, in 1998 the minority communities of Muslims, Sikhs, and
Scheduled Castes all voted at higher rates than did upper-caste Hindus. This has
not happened in every election, as the Muslim turnout dropped in one election
after the Gujarat massacre of 2002. But the basic pattern holds: those at the top of
india as a state-nation 81
the socioeconomic hierarchy are not at the top of voting turnout tables.58 Thus we
can conclude, from the tables just presented, that among previously marginalized
groups there is growing participation, a growing sense of efficacy, and a growing
commitment to Indian democracy as a way of managing diversity.
It is this spirit that has made for a smooth transition from the single-party rule
on the national level in the 1980s to the extraordinary proliferation of ‘‘centric-
regional’’ state parties that participate in the numerous state-nation ruling coali-
tions of the overall polity.59 Indeed, India has far more parties that participate in
ruling coalitions than any other democracy in the world. For example from 1945
to 1995, of the twenty-three OECD countries, the United States, Great Britain,
and New Zealand never had more than one party in the government. The greatest
number of parties that ever helped form a government at the center was eight, in
Belgium in 1945 and in France in 1947. In sharp contrast, in 2003, near the
completion of its five-year term, the BJP-led government had twenty-three parties
in the ruling coalition. The Congress-led ruling coalition in 2008 had seventeen
parties.60
As the above data indicate, the acceptance and use of this coalitional model is
not confined to the centrist parties like the Congress; it cuts across the left-right
divide as well as the ‘‘polity-wide’’ and ‘‘centric-regional’’ party divide. Even the
right-wing BJP, which has questioned the consensus on the secular character of
the Indian state, has accepted regional and linguistic diversity as one of the
foundations of Indian governance.
india as a state-nation:
past accomplishments and potential threats
58. Yogendra Yadav, ‘‘Understanding the Second Democratic Upsurge: Trends of Bahujan
Participation in Electoral Politics in the 1990s,’’ in Transforming India: Social and Political Dy-
namics of Democracy, ed. Francine R. Frankel, Zoya Hasan, Rajeev Bhargava, and Balveer Arora
(Delhi: Oxford University Press, 2000), tables 14 and 15. It should also be said that the reverse pattern
in India is of long standing. See Samuel Eldersveld and Bashiruddin Ahmed, Citizens and Politics:
Mass Political Behavior in India (Chicago: University of Chicago Press, 1978), table 14.5, p. 195.
59. This transition has been described by Yogendra Yadav and Suhas Palshikar in ‘‘From Hege-
mony to Convergence: Party System and Electoral Politics in the Indian States, 1952–2002,’’ Journal
of Indian Institute of Political Economy 15, no. 1/2 (January–June 2003).
60. For data on all countries except India, see J. Lane, D. McKay, and K. Newton, Political
Data Handbook: OECD Countries, 2nd ed. (Oxford: Oxford University Press, 1997), pp. 212–340.
82 crafting state-nations
61. See Linz and Stepan, eds., Breakdown of Democratic Regimes: Europe (Baltimore: Johns
Hopkins University Press, 1978), ix.
84 crafting state-nations
that boasted twenty-two other parties at the end of its five-year term in 2004.
Fourth, the anti-Muslim pogrom in February-March 2002 in Gujarat led to the
killing of approximately fifteen hundred people, about 75% Muslim, and was
followed by the electoral triumph of the BJP in the state under the leadership of
Narendra Modi, widely believed to be strongly complicitous with the pogrom.
The question after Gujarat was whether its model would be the future for
Indian politics and thus represent a fatal challenge to the state-nation model. After
the Gujarat elections, Ashutosh Varshney wrote the following, which we will
quote at length to give an indication of the worries that some important observers
had about the threats to pluralism and inclusiveness in India:
Clearly, if the Gujarat model became a dominant model in India, this would
bring about the sociopolitical destruction of India’s state-nation. We hope that this
will not occur, and we do not believe it is inevitable that it will.
In an article written in 2005, Linz, Stepan, and Yadav made the following
argument as to why we believed the spread of the Gujarat model was unlikely:
Let us briefly mention some reasons that make us more confident than many
others that the ‘‘Gujarat model’’ is not bound to be successful in India’s twenty-
62. Ashutosh Varshney, ‘‘Will the Stallion Baulk in Mid-Gallop?’’ Outlook, December 30, 2002.
india as a state-nation 85
seven other states. Gujarat has many features that make it exceptional in In-
dia. First, the Gujarat electoral model was aided by the fact that the BJP was
ruling in Gujarat without the actual, or at least potential, constraint of coali-
tional partners. . . . Second, Ashutosh Varshney presents data on deaths as
a result of communal violence per 1,000,000 of urban populations in seventeen
major states in India from 1950–1995. Gujarat, by far, had the highest death
rate.63 Third, of the seventeen states for which we have survey data on sup-
port for democracy, the state that had the highest number of explicitly anti-
democratic responses in 1998 was Gujarat. Fourth, Gujarat had emerged as the
safest electoral bastion for the BJP, and had witnessed the most intense Hindu-
fundamentalist mobilization of any state in the decade prior to the massacre.64
Fifth, the Godra incident in which 58 Hindus burned to death in a train return-
ing from Ayodhya, one of the main symbols for Hindu fundamentalism, helped
ignite the massacres. Without great complicity by incumbents, and unprece-
dented terrorism by civil society groups (whether Muslim or Hindu), a Godra
type incident will be an extremely rare occurrence. More generally, we can say
that the leaders of the BJP as a political party (who were in a governing coalition
with twenty-three partners) for reasons of parliamentary, coalitional, electoral,
and even very important national and international investment imperatives,
might well want to distance themselves from full association or complicity with,
the projects of such groups as the RSS, VHP and the Bajrang Dal. Finally, a
non-BJP government at the center, might not allow this, not only out of a
commitment to value India’s tradition of inclusiveness but also for reasons of
party competition. This would contribute to governability, a strong Indian state,
and wide-spread support for a state nation. Such a government at the center, in
all likelihood, would not tolerate an individual state leader’s incitement of a
Gujarat type anti-inclusionary campaign and its attendant massacres.65
Events since then have made us even more confident in this judgment. In fact,
far from the Gujarat pogrom ushering in a new triumphal period of Hindu
nationalism and BJP growth, it may have helped set into motion countervailing
powers, many of them coming from state-nation values and practices. At the time
of the Gujarat pogrom, the BJP was heading a coalition of over twenty parties at
the center, including some of the minority parties like the Akali Dal and the
63. Varshney, Ethnic Conflict and Civic Life (New Haven: Yale University Press, 2003), p. 97.
64. See Yogendra Yadav, ‘‘The Patterns and Lessons,’’ Frontline (India), January 3, 2003,
pp. 10–16.
65. Juan J. Linz, Alfred Stepan, Yogendra Yadav, ‘‘ ‘Nation State’ or ‘State Nation’? India in
Comparative Perspective,’’ in Democracy and Diversity, ed. Bajpai, pp. 104–105.
86 crafting state-nations
(1991)
100 (2009)
85
80 (1989)
60 Seats Won by BJP
40
20
2 2002
0 Gujarat
1984 1989 1994 1999 Pogrom 2004 2009
Year
Figure 2.3. Rise and Decline of BJP Seats in Lok Sabha, 1984–2009
Source: Election Commission of India.
National Conference. Only one coalitional ally, Shiv Sena, shared its exclusion-
ary and communal ideology, yet none registered an effective protest during or
immediately after the pogrom. Some of the self-corrective mechanisms of demo-
cratic political competition did unfold, but very slowly. Once the BJP-led coali-
tion, the NDA, lost the parliamentary election to the Congress-led UPA in the
2004 general elections, many of the BJP allies openly cited the Gujarat riots as a
reason for the loss of Muslim votes and the defeat of the coalition. This considera-
tion played an important role in the departure of some of the key NDA allies like
the Telugu Desam Party in Andhra Pradesh, the All India Anna Dravid Munetra
Kazhgam in Tamil Nadu, and the All India Trinamool Congress in West Bengal
between 2004 and 2009. The BJP paid the price for Gujarat, but not inside
Gujarat. The BJP’s own vote share declined from its peak of 25.6% in 1998 to
18.8% in 2009, slipping below the 20% mark for the first time since 1989.
The outcome of the parliamentary election held in 2009 underscored that the
BJP’s defeat in 2004 was not a one-time setback. Figure 2.3 shows that the BJP’s
electoral fortunes declined consistently following its peak in the late 1990s. From
the peak of 41.1% share of the national vote, the BJP-led alliance, the NDA,
slipped to 35.9% in 2004 and more sharply to just 24.1% in 2009. The election was
a serious setback for the BJP itself. Furthermore, not only did the BJP not succeed
india as a state-nation 87
in coming back to power, but its seats and votes also plummeted to the lowest level
since 1989, when its rise to power began. The BJP now faces the challenge of a
long-term erosion, since its social base is threatening to come apart.
An analysis of long-term trends in the sociology of voting by Yadav and his
coauthor, Suhas Palshikar, led them to conclude:
Thus the shift that was discernible in 2004 appears to have been completed in
2009. The NDA’s famous victory in 1999 represented the height of BJP’s success
in creating a ‘‘new social bloc.’’ The core of this bloc comprised voters from the
upper end of the social pyramid, a coalition of the privileged, which allowed the
party to draw heavily from this smaller pool of potential voters. . . . It required
ingenuous politics to convert ad hoc acquisitions of underprivileged into an
enduring political bloc. And it required statesman-like leadership to sustain a
coalition with allies who drew upon a very different social base that often in-
cluded the Muslims. In retrospect it seems that the BJP simply could not rise up
to this challenge. The story since 1999 is that of depolarization of this bloc:
the core bloc of the privileged has started drifting away from the BJP to the
Congress, demobilization appears to have set in for the social groups that had
been stapled and the supplementary groups have simply walked away with the
allies.66
It is important to remember that the elections of 2009 were held in the wake of
the attack in Mumbai by Pakistani terrorists in late November 2008 that killed at
least 170 people. The Indian response to the attack was a sign that many Indians—
Hindus and Muslims alike—even in the face of grave provocation still support
state-nation practices and values. Though the terrorists were Muslim, albeit from
Pakistan, there were no anti-Muslim riots, either in the immediate aftermath of
the attack in Mumbai or anywhere else in the rest of India. Muslim organizations
were quick to condemn the attack in unequivocal terms and even refused to allow
the burial of the perpetrators of the attack on the grounds that ‘‘people who
committed this heinous act cannot be Muslim.’’ The BJP did see this as an
opportunity to exert pressure on the government and tried hard to turn it into a
campaign issue for the parliamentary elections held a few months after the attack.
This issue had little appeal in the election. The BJP and its ultra-chauvinist ally
66. See Yogendra Yadav and Suhas Palshikar, ‘‘Between ‘Fortuna’ and ‘Virtu’: Explaining the
Congress’ Ambiguous Victory in 2009,’’ Economic and Political Weekly, vol. 44, September 26–
October 2, 2009, pp. 33–51. This entire 206-page special issue is devoted to articles on National
Election Study 2009 and is an invaluable source for analysts.
88 crafting state-nations
Shiv Sena won none of the six parliamentary seats in Mumbai itself, where the
attack took place.
It would be incorrect to infer from these findings that the BJP faces a terminal
decline. Its status as the second-largest national political party and potentially the
ruling party in a number of states does not seem to be threatened. Nor is there
compelling evidence to show that the Gujarat pogrom was the most decisive
factor that led to BJP’s electoral reversals. The complex political process does not
allow for such a mono-causal reading. But the electoral outcomes and political
developments since 2002 provide strong evidence to negate the worry that Gujarat
was going to be replicated across the country and thus be the beginning of the
end of the state-nation enterprise in India. While Modi is still chief minister of
Gujarat, as of this writing, in hindsight we know that fears of the ‘‘Gujarat model’’
being replicated all over the country were highly exaggerated. Indeed, instead of
becoming the launching pad for BJP’s rise to being a ‘‘natural party of gover-
nance,’’ Gujarat proved to be a turning point for the BJP’s political decline.
The anti-Muslim pogrom of 2002 in Gujarat reminds us that the success of
a state-nation is contingent on continuous political practices. Creating a state-
nation is not a one-shot affair but a continual effort. It also reminds us that what
is made can also be unmade. As in the case of nation-states, a state-nation is also
a politically imagined community that needs to be sustained through continu-
ous contestation and re-creation in the realm of ideas, institutions, and political
practices.
chapter three
International boundary
Line of Control undetermined 1972 Line of Control
State or union territory boundary
AFGHANISTAN
0 150 300 Kilometers
Jammu and 0 150 300 Miles
Kashmir
Punjab CHINA
PAKISTAN
Delhi
NEPAL
BHUTAN
Nagaland
BANGLADESH
Calcutta
(Kolkata) Mizoram
MYANMAR
Bombay
(Mumbai)
Bay of
Arabian Bengal
Sea
ANDAMAN
ISLANDS
(India)
Tamil
Nadu
LACCADIVE
ISLANDS
(India)
SRI NICOBAR
Laccadive LANKA ISLANDS
Sea (India)
MALDIVES
insurgent groups have been fighting for independence for sixty years, with no
victory or peace in sight. The presence of armed forces in the Valley of Kashmir,
the heart of the state of Jammu and Kashmir, is one of the most embarrassing facts
for any democrat in India. The stability of Jammu and Kashmir has been impeded
by the unfulfilled UN Security Council resolution for a plebiscite to determine its
status vis-à-vis conflicting Indian and Pakistani claims of sovereignty, long periods
of formal and informal central government rule by Delhi, insurgencies, popular
four indian cases that challenge state-nation theory? 91
protests, and three wars between India and Pakistan. As the reader can see in
figure 3.1, these four cases involve four states on the periphery of India.
This short chapter, of course, cannot do justice to the historical and socio-
logical complexities of these cases. Our goal is much more limited. What do these
cases tell us—or not tell us—about the limits of our state-nation model?
The Punjab crisis presents us with the first challenge for our reading of India’s
democratic experience. Here the crisis refers to a series of political developments
in the state of Punjab in the 1980s and 1990s that brought the constitutional order
within the state to a standstill, gave rise to a powerful and violent secessionist
movement, and complicated the relationship of the Sikh minority with the Indian
union both inside and outside the state of the Punjab. The resolution of the crisis
after 1996, as sudden and surprising as its onset in the 1980s, raises two sets of
largely unanswered questions about what caused the disequilibrium and what led
to the reequilibrium. First, did the Punjab crisis represent a failure of the state-
nation model to accommodate diversity and democracy? How intractable were
relations between Sikhs and the Indian political system in the late 1970s? Were
state-nation policies increasingly being rejected by the 1970s? Second, was the
reequilibrium after 1996 only the result of the effective application of the over-
whelming coercive resources of the Indian state, a state to which the Sikhs are not
reconciled? If the answer to one or both of these questions is yes, major reserva-
tions about the state-nation model would seem legitimate. Let us find out.
By 1980 the Sikh community, a religious minority comprising less than 2% of the
country’s population but nearly a two-thirds majority within the state of the Punjab,
had in many ways done well in independent India: Sikhs were economically better
off than the rest of India, were educationally more advanced, enjoyed historical
overrepresentation in the armed forces, and had a fair political representation.1
1. Indeed, Paul R. Brass, in his authoritative article written during the height of the Punjab
crisis, argues that Sikh economic resentment toward the policies of the Indian federation was hard to
see as a full explanation for the secessionist movement, because legitimate grievances of a federal
nature concerning the location of the capital and the division of river waters notwithstanding, ‘‘on
the great majority of aggregate economic indicators, Punjab is at the top of or very close to the top in
comparison with all other Indian states. Moreover, per capita plan expenditures have been consis-
tently higher in Punjab than in every other major Indian state for the last twenty years.’’ See Brass,
‘‘Socioeconomic Aspects of the Punjab Crisis,’’ in his Ethnicity and Nationalism: Theory and Com-
parison (London: Sage, 1991), p. 229.
92 crafting state-nations
2. The text of the Anandpur Sahib Resolution is reprinted as appendix 1 in Devinder Singh,
Akali Politics in Punjab (1964–1985) (New Delhi: Anupreet Marwah Publishers, 1993), pp. 225–232.
The above quotation is from Resolution One, p. 235.
3. Harish Puri, Paramjit Singh Judge, Jagrup Singh Sekhon, Terrorism in Punjab: Understand-
ing Grassroots Reality (New Delhi, India: Har-Anand Publications, 1999), pp. 37–38, quotation from
p. 8.
4. See the excellent chapter ‘‘The Punjab Crisis and the Unity of India’’ in his Ethnicity and
Nationalism, quotation p. 210.
94 crafting state-nations
moderate form of Sikh politics, they assert that ‘‘the scales were, however, tilted
against the Akali Dal after the return of Indira Gandhi’s Congress to power at the
Centre in 1980 [and] the immediate dismissal of the Badal government in Punjab
(by a blatant abuse of Article 356).’’5 Direct rule by the center and the dismissal
of a democratic majority government in a state with a minority population in
themselves erode the multiple but complementary identities and state-nation
sentiments.
Worse, there is substantial evidence that Indira Gandhi not only dismissed the
moderate Akali government but also extended covert support to the extremist
leadership represented by Sikh militant Sant Bhindranwale so as to divide and
discredit moderate Sikh leaders and to prepare the way for political dominance by
her Congress Party against the faction-ridden Akali Dal. Brass argues that ‘‘the
involvement of Sanjay Gandhi in the recruitment of Bhindranwale also meant
that that criminal actions, manipulation of the police and the judiciary, and the
use of violence were considered acceptable tactics by Congressmen, by the po-
lice, and by its allies to defeat and discredit the Akali Dal.’’6
This is when the Punjab crisis erupted. Dislodged from power in 1980, the
Akali Dal launched an agitation that combined routine regional demands like
more power to states within the Indian union and a greater share of river water
distribution with some religious demands of the Sikh community. Sensing an
opportunity to disable their opponents for a long time to come, the Congress
leadership actively encouraged a militant faction within Sikh politics so as to
displace the moderate leadership of the Akali Dal. The game initiated by the
ruling party soon took on a life of its own, and the militant faction took to arms.
Thus ensued a deadly game of competitive extremism in Sikh politics, resulting in
the rise of armed militant groups, many under the influence of Bhindranwale,
who managed to take control of the most sacred Sikh holy shrine, the Golden
Temple in Amritsar. The government of India responded in June 1984 with ‘‘Oper-
ation Bluestar,’’ in which the army entered the Golden temple with tanks and
eliminated the militants. This action was perceived by ordinary Sikhs as an attack
on their faith. In a retaliatory action, the Sikh bodyguards of Prime Minister Indira
Gandhi assassinated her in November 1984. The next few days saw a massacre of
thousands of Sikhs in Delhi and some other cities, with the connivance of police
and state machinery at that time and an absence of any action against the guilty
thereafter.7
This turn of events spawned a full-blown militancy. Ordinary Sikhs felt deeply
alienated, and several militant and terrorist groups sprang up to demand the
creation of Khalistan, which would be a Sikh homeland. The Khalistan move-
ment did not enjoy great popular support but was powerful enough to split Akali
Dal and sideline the moderate Sikh leadership, disrupt regular constitutional
government, and render worthless a major agreement between the government of
India and Sant Longowal, an important moderate Akali Dal leader, in 1985.
Attempts to lodge a popularly elected Akali Dal government failed in the face of
rising political violence in the state by the armed militants. The next five years saw
the breakdown of constitutional order, an imposition of central rule in the state,
large-scale terrorist violence, and an increasingly violent and undemocratic re-
sponse by the armed forces.
As the battle between the militants and the state reached a stalemate, a Con-
gress government was put in place in 1992 through a dubious election that was
boycotted by the Akali Dal and a vast majority of Sikhs.8 This ‘‘popular’’ govern-
ment presided over some very heavy repression by the security forces, including a
large number of extrajudicial killings of suspected militants. The militant move-
ment could not stand this repression.9
This is when the Congress lost power at the center in the parliamentary elec-
tions held in 1996. The new political context encouraged the moderate Sikh
leaders to participate in the next state assembly elections held in 1997, which
produced an Akali-led government. By 1998 Punjabi politics seemed to have
reequilibrated. At the height of the crisis, there were a thousand or more violent
deaths every year for several years in a row. But from 1998 to 2008, according to
figure 3.2, only thirty-one deaths occurred; indeed in seven of those years, there
were zero violent deaths.
7. For a comprehensive documentation of the legal records concerning the massacre and the
court proceedings, see Manoj Mitta and H. S. Phoolka, When a Tree Shook Delhi: The 1984 Carnage
and Its Aftermath (New Delhi: Roli Books, 2007).
8. An analysis of the electoral turnout in this election by Yogendra Yadav confirmed that there
was a direct and negative relationship between electoral turnout and the proportion of non-Dalit
Sikhs among the electorate. See Yogendra Yadav, ‘‘Who Won in Punjab: Of the Real Contest,’’
Frontline (India), vol. 10, April 1992, pp. 122–126.
9. This was meticulously documented by Ram Narayan Kumar, Amrik Singh, Ashok Agrwaal,
and Jaskaran Kaur in Reduced to Ashes: The Insurgency and Human Rights in Punjab (New Delhi:
South Asia Forum for Human Rights, 2003), available at www.punjabjustice.org. Interestingly, the
Akali Dal–led government that came to power after reequilibriation in the state refused to order a
fair investigation into these gross violations of human rights under the Congress regime, much to the
chagrin of human rights groups.
96 crafting state-nations
3000
Civilians
2500 Terrorists
Security Forces
2000
1500
1000
500
0
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
The rapidity of the reequilibrium took everyone by surprise, just as the onset of
crisis had about fifteen years earlier. This brings us to another of the major
questions about the Punjab crisis. Why such rapid equilibrium? Was it a pyrrhic
victory for Indian democracy? Certainly a crucial factor for the defeat of the
Khalistan movement in the Punjab was India’s useable state coercive apparatus.
As we have argued, for most would-be separatists in India, in sharp contrast to the
situation in the U.S.S.R. or Yugoslavia, the loyalty of the security forces to the
central state and the state-nation is a given, as is the certainty that force will be met
with greater force. Geopolitically speaking, there is no exit. The end of insurgency
in the Punjab was no doubt made possible by massive, undemocratic, and often
unconstitutional use of state repression. But the use of coercive apparatus suc-
ceeded in the Punjab in a way it did not in states like Nagaland, because the
Indian state and its institutions enjoyed a certain legitimacy in the eyes of the
people of the Punjab. This is why the Khalistan movement never enjoyed wide-
spread popular support among the Sikhs. This is what made a return to normal
politics so quick and complete.
The insurgents may have been defeated, but at what cost to the legitimacy of
the Indian state? Do Sikhs believe that the methods used by the armed separatists
were legitimate and that the tactics used by the Indian state, especially the Punjab
police, were illegitimate? If so, there could still be a serious problem of political
identities in the Punjab for India. The base for an insurgency, or at least a series of
four indian cases that challenge state-nation theory? 97
table 3.1
Rejection of Pro-Khalistan Violence by the Sikh Community: Attitudes toward Methods Used
by the Militants and Punjab Police in the Punjab by Religion, 1997
Opinion about the
Opinion about the methods used by Punjab
methods used by the Pro- police to suppress pro-
Khalistan militants Khalistan militants
Response Sikh Hindu Sikh Hindu
‘‘Justified’’ 5% 4% 27% 41%
‘‘Somewhat justified’’ 21 12 33 36
‘‘Not justified’’ 66 77 30 14
‘‘Can’t say/do not know’’ 8 7 10 10
(N) (3,138) (1,628) (3,138) (1,628)
Source: CSDS Data Unit, Punjab State Assembly Elections Study, Exit Poll, 1997.
Note: Responses to the following questions: ‘‘Do you believe that the means used by the Kharku [self-
description of the pro-Khalistan militants] to fulfill their objectives were justified or not?’’ and ‘‘Do you
believe that the means used by the Punjab Police to deal with the Kharku were justified or not?’’
Four years after the poll referred to in this table, a similar battery of questions was asked after the
Punjab State Assembly elections of 2002. The same pattern of responses was repeated—except for an
even stronger rejection of the methods used by the pro-Khalistan militants. In the 1998 Indian NES
‘‘post-poll study,’’ more respondents in Punjab than in any other state said that life and property were
safer now than five years ago.
Percentages may not add up to 100 due to rounding.
disloyal violent activities that may weaken democracy in the future, might still
be there.
Fortunately, we can begin to answer this fundamental question due to an
unusually large exit poll with a sample of 4,950 randomly selected respondents for
the Punjab State Assembly Elections of 1997. In answer to the question ‘‘Do you
believe that the means used by the Kharku [a word coined by the Khalistan
militants to describe themselves] to fulfill their objectives were justified or not?’’
for every Sikh who answered ‘‘justified’’ (5%) there were more than twelve (66%)
among the Sikhs—whose independence was being sought, remember—who an-
swered ‘‘not justified.’’ Concerning the appropriateness of the (often quite violent)
methods of the Punjab police to counter the Kharku, 27% of Sikhs said that the
methods used by the police were ‘‘justified,’’ 33% said ‘‘somewhat justified,’’ and
only 30% answered ‘‘not justified’’ (see table 3.1).
Was military victory alone the key to the reequilibration? Do Sikhs deeply
distrust the center despite the ‘‘peace’’? Or did the reestablishment of normal
politics in the context of the unimpeded operation of state-nation mechanisms
contribute to a democratic reequilibration?
98 crafting state-nations
10. Various surveys by the CSDS show that Congress assembly election vote share among the Jat
Sikhs, the community at the heart of Sikh militancy, went up steadily from 16% in 1997 to 23% in
2002 to 31% in 2007. The gap between the Akali and the Congress vote share among this community
has steadily come down from 49 points in 1997 to only 24 points in 2007, thus indicating a decline in
social polarization. Full details of these and other CSDS surveys are available at www.csds.in/
questionnairebank.html#NES.
four indian cases that challenge state-nation theory? 99
Election Study revealed that 63% of those polled in the Punjab expressed a ‘‘great
deal’’ or ‘‘some’’ trust in the central government, which, as the reader will remem-
ber, is substantially above the average level of trust in central government among
the eleven longstanding federations documented in table 2.9.
To conclude our discussion: Was the Punjab crisis evidence of the failure of
the state-nation model in India? The evidence suggests that state-nation politics
were functioning reasonably well until 1980. The Punjab crisis was akin to a
‘‘democratic breakdown’’ familiar to students of comparative democratic theory,
except that it did not involve the entire system. This sub-system breakdown, like
most democratic breakdowns, was neither inevitable nor irreversible.11 This was
one of the many instances in which a democratic regime subverts its own system.
In the Punjab democratic incumbents at the center contributed to the collapse of
the emerging state-nation democratic political game, which led to the loss of
popular legitimacy in the Punjab. The cause of this state-induced crisis must be
sought in politics itself, specifically in the failure of the political elite at the center
to allow a moderate minority to rule peacefully. The breakdown in the Punjab was
not due to the inability of state-nation policies to manage diversity democratically
but rather to the central government’s gross violation of state-nation policies.
Indeed, in our introduction to the theory of state-nation, we argued that a funda-
mental principle necessary for the smooth functioning of a state-nation is what the
Germans call Bundestreue (loyalty to the federation). The prime minister of India,
Indira Gandhi, by her policies towards the Punjab, deeply violated Bundestreue.
The defeat of the insurgency was due to the overwhelming use of force by the
center. But the democratic reequilibration we have just documented was made
possible by a change in political context. State-nation politics were allowed to
generate governments and policies, and state-nation principles and practices were
re-embraced by former enemies with renewed commitment. The fifteen years of
the Punjab crisis had taught the political leadership on both sides lessons in
moderation: the tired Akali leaders were acutely aware of the costs of remaining
outside the system, while the new generation of Congress leadership was less
interventionist in non-Congress-led states. More important, the political climate
in this ‘‘post-Congress polity’’ was more conducive to power-sharing. The Supreme
Court had stepped in to prevent the frequent misuse of the emergency provisions
of the constitution by the central government to dismiss inconvenient state govern-
ments. Congress was no longer the natural party of governance at the center—there
11. See Linz and Stepan, eds., The Breakdown of Democratic Regimes (Baltimore: Johns Hop-
kins University Press, 1978), preface, esp. p. ix.
100 crafting state-nations
was enough room in the new coalition arrangements for everyone. The Akalis
entered into a stable coalition with the BJP both at the center and the state. Instead
of the danger of permanently being in opposition, the Akali Dal is now quite
regularly in power, either in the state or at the center. The critical elections of 1996,
1997, 2002, and 2007 were largely devoid of ‘‘ethnic outbidding’’; indeed, Akali and
Congress politicians alike frequently spoke of the need for ‘‘peace outbidding.’’
The history of separatist struggles in the two northeastern states of Mizoram and
Nagaland offers an instructive contrast about the success of and limits to the state-
nation policy followed in India.12 Secessionism in Mizoram came to an end with
an accord in 1986; today, the state is arguably the most peaceful state in the
northeast. However, Naga insurgency has continued for more than sixty years.
The contrast invites us to think about the conditions that led to the rise of the
insurgency in both of these states and the factors that led to the resolution in
Mizoram but not in Nagaland. Were these insurgencies linked to the failure of the
state-nation model? Did this model contribute to the resolution in Mizoram? If
so, why did it not work in Nagaland?
Mizoram and Nagaland are examples of politically activated, territorially con-
centrated sociocultural groupings of people that would seem extremely difficult to
reconcile with membership in a nation-state community if the prevailing criterion
was a common culture with India. Less than 3% of the population in Nagaland
and less than 2% in Mizoram speak Hindi, the most common language in India.
The population in both the states is predominantly Christian—87% in Mizoram
and 90% in Nagaland. The Hindus, who comprise 81% of the rest of India’s
population, are a microscopic minority in both of these states. Scheduled Tribes
in the rest of India represent only 8% of the population—but 95% of the popula-
tion in Mizoram and 88% in Nagaland.13
Earlier in chapter 2 we documented that most Indians shared some civiliza-
tional identities despite significant cultural differences. Further, most parts of Brit-
ish India developed, via their involvement with the Congress-led Independence
12. For a detailed listing of all the insurgent groups operating in India’s northeast, see South
Asian Terrorism Portal at www.satp.org.
13. All demographic details are from Census of India, 2001, Office of The Registrar General
of Census. Available at www.censusindia.gov.in/CensuseDatae2001/CensuseDataeOnline/, ac-
cessed July 13, 2009.
four indian cases that challenge state-nation theory? 101
Mizoram
14. For an insider’s account of Naga history, see M. Alemchiba, A Brief Historical Account of
Nagaland (Kohima, India: Naga Institute of Culture, 1970). On Mizoram, see A. G. McCall, Lushei
Chrysalis (London: Luzac, 1949).
15. Nehru acknowledged this in the case of Nagaland in a confidential letter to the chief
minister of Assam: ‘‘I feel that we have not dealt with this question of the Nagas with wisdom in the
past. We must not judge them as we would others who are undoubtedly part of India. The Nagas
have no such background or sensation and we have to create that sensation among them by our
goodwill and treatment.’’ Secret and Personal Letter dated May 13, 1956, to B. R. Medhi, chief
minister, Assam. Reproduced in Sanjoy Hazarika, Strangers of the Mist: Tales of War and Peace from
India’s Northeast (London: Penguin, 1994), p. 360.
102 crafting state-nations
traditional practices and norms, the drafters of the Indian constitution created
from the beginning the possibility that some indigenous cultures within certain
federal states would be allowed to make some special laws to protect their culture,
such as not allowing nonmembers of the indigenous group to vote in local elec-
tions or to purchase land. The drafters were thus alive to the challenge of radical
difference posed by areas such as Mizoram. The Constituent Assembly appointed
a special subcommittee on the Tribal and Excluded Areas of the North East
Frontier headed by G. N. Bordoloi, a Congress member from Assam who had
empathy for and knowledge of the tribes in the northeast. This subcommittee
drafted the special provisions of the Sixth Schedule of the constitution. As per the
recommendations of the subcommittee, duly incorporated in the constitution,
Mizoram was given the status of an Autonomous District within the state of
Assam. This status granted a state a good deal of leeway in determining its own
practices, especially in dealing with ‘‘customary’’ matters under the partially dem-
ocratically elected district council.
Nonetheless, on Indian Independence Day, there was still much confusion
about whether Mizos were free to join Pakistan or Burma—as well as ambivalence
about what kind of relationship they would have with India if they became part of
India. A resolution passed by a meeting of Lushei Hills (the older name for today’s
Mizoram) chieftains and other prominent citizens under the chairmanship of the
district superintendent on August 14, 1947, urged the governor of Assam to inform
them ‘‘whether the Lusheis are at this stage allowed the option of joining any other
Dominion, i.e., Pakistan or Burma.’’ In the event that they were to enter the
Indian Union, the resolution demanded that ‘‘the Lusheis will be allowed to opt
out of the Indian Union when they wish to do so subject to a minimum period of
ten years.’’16
The turning point in the relationship between Mizoram and India came with
the great famine of 1958–59 and the failure of the state of Assam and the govern-
ment of India to respond properly.17 The Mizo National Famine Front, initially a
nonpolitical body formed in 1959 to provide relief and organize public action, was
converted in 1962 into the Mizo National Front, a political party, as Pu Laldenga,
a former government employee, rose to become its leader. After an inauspicious
beginning in the 1962 elections, the MNF tasted success in the assembly by-
16. Cited in B. B. Goswami, The Mizo Unrest: A Study of Politicisation of Culture (Jaipur:
Aalekh Publishers, 1979), pp. 134–135.
17. For a detailed account and critique of the official response first to the famine and then to the
MNF-led insurgency in Mizoran, see Amritha Rangasami, ‘‘Mizoram: Tragedy of Our Own Mak-
ing,’’ Economic and Political Weekly 13, no. 15, April 15, 1978, pp. 653–662.
four indian cases that challenge state-nation theory? 103
elections of 1963. Then followed a rapid turn of events: MNF demanded the
dismissal of the district council leadership but did not succeed; a split within the
MNF put pressure on the leadership to go for an openly secessionist plank, and
the Assam Congress leaders tried to play the MNF against its electoral opponent
for district council leadership, the Mizoram Union.18 By the end of 1965, this
sequence of events culminated in the MNF declaring the formation of a parallel
government under the leadership of Pu Laldenga.
The Mizo insurgency began in 1965. With arms largely obtained in neighbor-
ing East Pakistan, ‘‘Laldenga stuck simultaneously at several places with Opera-
tion Jerico. Well-trained guerrillas overran Aizawl [the administrative center of
Mizoram]: the Treasury, the radio station and the police station fell into their
hands. Lungleh, the other major town was also swamped and the guerrillas had
the run of the district.’’19 The Indian Defense Ministry responded massively: ‘‘Air
raids were ordered for the first time in free India’s history’’ against domestic
opponents.20 Taking a page out of the U.S. ‘‘strategic hamlet’’ counterinsurgency
tactics in Vietnam, the Home Ministry in Delhi ‘‘ordered the regrouping of
villages into virtual concentration camps by security troops.’’21 The Mizo National
Army, the militant wing of the MNF, used guerrilla tactics in the hilly terrain to
hold Indian armed forces but could not make dramatic advances, especially after
it lost its secure base and source of supply in neighboring East Pakistan, when,
with great help from India, East Pakistan seceded from Pakistan in 1971.22 In 1975
and 1977 some key MNF fighters surrendered.
This was the time when formal negotiations with the government of India
began but did not succeed. The legitimate political wing of the MNF split, yet the
more unified underground organization controlled by Laldenga was still powerful
enough to extract ‘‘taxes’’ and organize guerrilla raids. However, key actors in the
MNF and the Indian government increasingly saw that the situation was what is
now called in the literature a ‘‘mutually hurting stalemate’’—that is, a situation in
which neither side can be completely beaten, nor completely win.23
18. This period between the formation of MNF and the formation of parallel government in
Mizoram’s political history needs more research, for this could be another case of ‘‘democratic
breakdown’’ within a sub-system. This account draws from evidence in Goswani (n. 16), pp. 139–156.
19. Hazarika, Strangers of the Mist, p. 113.
20. Ibid.
21. Ibid.
22. Ibid, p. 117.
23. See, for example, William I. Zartman, ‘‘The Timing of Peace Initiatives: Hurting Stalemates
and Ripe Moments,’’ Global Review of Ethnopolitics 1, no. 1 (2001), pp. 8–18. In chapter 7, where we
offer a new theory of federacy, we document how a ‘‘hurting stalemate’’ played an important role in
the 2005 federacy solution that ended a civil war in Aceh, Indonesia.
104 crafting state-nations
How and why did this ‘‘mutually hurting stalemate’’ end? Why did an enduring
peace emerge? The historic Mizoram Accord of 1986 utilized almost every pos-
sible state-nation policy available and created some new ones. Both the Mizos and
India fully accept and honor this agreement.
The long history of armed insurgency and state repression was brought to an
end in the Mizoram Accord between the MNF, represented by Laldenga, and the
government of India on June 30, 1986.24 Negotiations between the two parties took
well over a decade. There was an earlier accord in 1975 that involved a smaller
faction of the MNF. Yet Laldenga and his band of loyal supporters remained
underground. The government of India agreed to give Mizoram the status of a full-
fledged state with a special provision to safeguard its autonomy. This special
provision, later incorporated in the constitution as Article 371(g), stated: ‘‘Notwith-
standing anything contained in the Constitution, no act of Parliament in respect of
(a) Religion or Social practices of the Mizos, (b) Mizo customary Law or pro-
cedure, (c) Administration of Civil and Criminal Justice involving decisions ac-
cording to Mizo customary Law, (d) Ownership and transfer of land, shall apply to
the State of Mizoram unless the Legislative Assembly of Mizoram by a resolution
so decides.’’ The center also agreed not to amend or repeal ‘‘inner line’’ regulation
without consulting the Mizoram government. There was also a co-signed written
agreement five days before the accord between the Congress Party and the MNF
that Congress, then in power in the state, would allow Laldenga to take over as
interim chief minister pending a regular election.25 For at least two years before
this Mizoram Accord, the major civil society groups in Mizoram, especially the
church groups, had been pressing Congress, MNF, and the government of India to
reach a peace settlement. Thus, the two major groups in political society and the
most important group in civil society, the church, were united behind ending the
insurgency and crafting an enduring peace. The government of India did not
concede any ground on the demand for unification of ‘‘Mizo-inhabited areas’’ in
other states like Manipur, and the MNF agreed to shelve this issue.
In return, the MNF agreed to cease hostilities, to ‘‘bring out all the under-
ground personnel of MNF with their arms, ammunitions and equipments,’’ to
stop supporting other insurgent groups in the northeast, to amend its own consti-
tution in accordance with the constitution of India, and to drop the demand for a
separate sovereign state of Mizoram.
The Mizoram Accord was followed by an institutionalization of the routines of
democratic politics—with surprising success. Within a few months of the accord,
the necessary constitutional amendments were enacted. The incumbent chief
minister from Congress, true to his word, stepped down to make way for Laldenga
to become the interim chief minister. In 1987 Mizoram became a full state, and
the MNF won the first election to the state legislative assembly. The MNF gov-
ernment did not survive long and lost to Congress in the midterm election of 1989.
Since then, the state has had regular elections and transfers of power: Congress
won in 1993; the MNF came back to power in 1998 and retained it in 2003 but lost
to Congress again in 2008. Mizoram is one of the most peaceful places in the
northeast, notwithstanding small insurgency groups representing the non-Mizo
minorities within Mizoram. Laldenga passed away in 1980 and was succeeded as
chief minister by his underground colleague Zoramthanga, once the number two
in the underground chain of command, who headed two MNF governments and
helped some of the neighboring states negotiate with their insurgents.
In an interview with Stepan, Zoramthanga recited, virtually from memory, key
passages from the Mizo Accord and the Sixth Schedule. Zoramthanga remarked
to Stepan that in effect, Mizoram had all of the benefits of being an independent
state and all of the advantages of being a well-treated and well-subsidized member
of the Indian federation.26
Zoramthanga’s opinion is widely shared by his fellow Mizo citizens. CSDS
carried out a survey of 1,116 randomly selected electors following the state assem-
bly elections in Mizoram in 2003. The survey findings indicate a surprisingly high
level of integration into India’s federal democracy: as many as 84% expressed a
preference for democracy. Despite the historical lack of a strong link to Indian
culture or history, a majority of the respondents included India as constitutive of
their identity: only 32% of the respondents identified themselves as ‘‘only Mizo.’’
Two-thirds of those who had an opinion supported the Mizoram Accord of 1986,
and 76% of those polled supported the special autonomy measures by which only
Mizos can vote in local elections and only Mizos can buy land in Mizoram.27
But why does the insurgency in Nagaland continue? An analysis of the situa-
26. Interview of Stepan with Chief Minister Zoramthanga, Aizawl, January 2003.
27. Mizoram Assembly Election Study 2003, CSDS Data Unit, Center for the Study of De-
veloping Societies, Delhi, India, 2003. Stepan also carried out interviews with both Congress and
MNF political leaders in Mizoram in January 2003 and arrived at similar conclusions.
106 crafting state-nations
tion there can tell us much about the conditions in which state-nation policies
can, and cannot, work. Before we turn to Nagaland, let us summarize the key
points from the Mizoram story.
Insurgency and peace are relational concepts. There were five key relation-
ships of Mizoram to the federal government of India involved in the MNF insur-
gency and the eventual peace. First, there was weak Mizo involvement and identi-
fication with the dominant religions and languages of India, together with almost
no participation in the politics of pre-British India, British India, or the Congress-
led Independence movement. Second, in independent India, major Mizo leaders
and many of their followers came to see themselves as a nation that was not being
treated well by the government of India and launched insurgency activities for
complete independence. Third, the military conflict would have ended if either
the government of India or the insurgents had won a decisive victory at any point.
Neither outcome happened, so the ‘‘mutually hurting stalemate’’ ensued. Fourth,
the government of India made an innovative offer for Mizoram to have a form of
extreme asymmetrical federalism with large guarantees of cultural autonomy.
Fifth, civil society and political society in Mizoram, led by a united armed under-
ground leader, was sufficiently united to be able to arrive at a self-binding accep-
tance of the offer; normal state-nation politics rapidly ensued, and peace has
endured.
Nagaland
For all intents and purposes, the relationship between Nagaland and India was
very similar on the first four factors.28 However, Nagaland was—and remains—
extremely dissimilar on the fifth factor, so the mutually hurting stalemate con-
tinues, despite periodic ceasefires since 1997.
28. Five valuable sources for Nagaland with some Mizoram comparisons, particularly on the
unity or disunity of the insurgent movements and its effect on peace, are: Sanjib Baruah, ‘‘Confront-
ing Constructionism: Ending India’s Naga War,’’ Journal of Peace Research 40, no. 3 (2003), pp. 321–
338; H. Srikanth and C. J. Thomas, ‘‘Naga Resistance Movement and the Peace Process in Northeast
India,’’ Peace and Democracy in South Asia 1, no. 2 (2005), pp. 57–87; Leanne C. Tyler, ‘‘Common
Origins, Divergent Outcomes: A Comparative Analysis of India’s Nagaland and Mizoram Wars of
Secession,’’ unpublished paper written at the Department of Political Science, Columbia University,
May 8, 2009; the chapters on Nagaland and Mizoram in Ved Marwah, Uncivil Wars: Pathology of Ter-
rorism in India (New Delhi: Harper Collins, 1995), pp. 224–286; and Hazarika, Strangers of the Mist.
The most comprehensive and sympathatic, if dated, account of the violations of human rights in
Nagaland is Luingam Luithi and Nandita Haksar, Nagaland File: A Question of Human Rights (New
Delhi: Lancer, 1986). For a fairly detailed account and critique of the official response first to the
famine and then to the MNF-led insurgency in Mizoram, see Amritha Rangasami, ‘‘Mizoram: Trag-
edy of Our Own Making,’’ Economic and Political Weekly, vol. 13, no. 15, April 15, 1978, pp. 653–662.
four indian cases that challenge state-nation theory? 107
Why the difference between successful accord in Mizoram and the continuing
stalemate in Nagaland? The main answer is that the Naga leadership has never
produced a united insurgent organization. Even dismissing smaller groups, there
have always been between anywhere from two to four organizations speaking for
the Nagas. In 2005 there were ‘‘four Naga militant groups—NNC (Adino), NNC
(Panger), NSCN (IM) and NSCN (K).’’29 The NSCN (Isaac-Muivah) faction is
the largest group, drawing some of its leaders from Tangkhul Naga tribe, largely
based in the state of Manipur. This group is engaged in an internecine war with
NSCM (Khaplang), led by Konya Nagas and mainly based in Myanmar. Another
group, NSCN (Unification), has also become active since 2007. What this means
is that whenever some organizations enter into peace talks with the government of
India, a process of ‘‘outbidding’’ rapidly occurs. This occurred around Indepen-
dence, when ‘‘moderate leaders like Aliba Imti and T. Sakhrie who were negotiat-
ing basically for greater autonomy within the Indian Union whereas leaders like
Phizo gave the clarion call for independence.’’30 In 1975 the Shillong Accord was
signed by breakaway groups but this led to even greater division and conflict
among the Nagas.31 Inevitably, the insurgent group interested in peace is de-
nounced by the others. In the language of international relations peace theory,
Nagaland has multiple militarily credible ‘‘spoilers’’ to any negotiation. This con-
trasts sharply with the single united negotiator and the absence of ‘‘spoilers’’ in
Mizoram.32
Why the Nagas are divided needs greater research, but some simple facts
stand out: In 2001, 73% of the population in Mizoram spoke the Mizo language,
whereas only 13% of the Nagas spoke Ao, the most common language in the
state.33 Furthermore, all of the Mizo military leaders came from the Lushei Hills,
so their claims for a ‘‘greater Mizoram’’ to include all the Mizos in neighboring
states were bargainable with the government of India.34 In contrast, the home and
base of the most important leader of NSCN (I-M), Th. Muivah, is the neighbor-
35. The National Election Study 2004 by the CSDS included a sample of 522 randomly chosen
respondents from Nagaland and was perhaps the first-ever survey of political opinions and attitudes
in the state. The survey brought out an unwillingness to accept the Indian identity: about 57% of
respondents identified themselves as ‘‘only Nagas’’ as againt 9% ‘‘only Indian’’ and 34 % ‘‘Indian and
Naga.’’ At the same time, the alienation did not affect attitudes to the peace process: 85% had heard
about the negotiations, and two-thirds of those who had an opinion were optimistic of a positive
outcome. The survey also revealed a lack of strong support for the hardline agenda of ‘‘greater
Nagaland’’ that has blocked the negotiations.
36. For an excellent social science and political discussion of this issue, see Baruah, ‘‘Confront-
ing Constructionism,’’ esp. p. 333.
four indian cases that challenge state-nation theory? 109
In the preface to this book we argued, in effect, that relatively secure ‘‘stateness’’ is
crucial for either a democratic nation-state or a democratic state-nation and that
both ‘‘must be states if they are to work.’’ In this fundamental sense, part of the,
to date, unsolvable problem about Jammu and Kashmir is that its stateness has
never been internationally recognized, nor peacefully settled between India and
Pakistan.
The state of Jammu and Kashmir does not, by itself, present diversities that are
any deeper or more challenging than, say, those of Mizoram and Nagaland.
Geography, culture, and history tie this region to the rest of India in ways that are
deep and well recognized.37 A complex pattern of cultural ties yet distinctiveness
vis-à-vis the rest of India combined with deep yet interconnected internal differ-
ences do not, by themselves, make Jammu and Kashmir a case apart from the rest
of India, nor do these preexisting conditions, by themselves, account for the
Kashmir problem as it stands today.
What makes the problem so very complicated is the complex political history
of the state in the last century. Before India’s independence, Jammu and Kashmir
was a princely state with a Hindu ruler and an overwhelmingly Muslim popula-
tion. The opposition to the princely rule was led by Sheikh Abdullah, a charis-
matic Kashmiri leader and a personal friend of Jawaharlal Nehru who trans-
formed the Jammu and Kashmir Muslim Conference into a secular Jammu and
Kashmir National Conference, an ally of the Indian National Congress.
In the Partition of India in 1947, Jammu and Kashmir emerged as an inter-
nationally disputed territory between India and Pakistan. The principle of acces-
sion in the more than five hundred principalities was that the princely ruler was
free to decide whether to join India or Pakistan. Presumably, when a state had a
Muslim majority and bordered Pakistan it would join that state. However, the
37. Jammu and Kashmir comprises three distinct regions—Kashmir, Jammu, and Ladakh—
each with very different socio-religious composition. The Kashmir Valley, the largest region at the
time of India’s partition and at the heart of political alienation, is now almost entirely Muslim and
Kashmiri speaking. Over the last three decades, the population of Kashmiri Hindus has come down
from about 5% to just 2%. Kashmiri language and Kashmiri cultural identity binds the Muslims and
the Hindus in the valley. A predominantly Sufi Islam distinguishes the form of Islam practiced in
Kashmir from the rest of the subcontinent and outside. The Jammu region—now as populous as
Kashmir—is majority Hindu and speaks many languages and dialects including Hindi. Ladakh is the
third and the least known region in the state. A mountainous and scantily populated region with vast
territory bordering China whose population at independence was largely Buddhist, Ladakh is now
evenly split between Muslims and Buddhists. The region has its own languages, history, and culture.
110 crafting state-nations
ruling prince of Jammu and Kashmir, Maharaja Karan Singh, was a Hindu from
Jammu and equivocated. But when insurgents with the support of Pakistan threat-
ened his rule, he asked for Indian military support and agreed to accede to India in
return. It is important that Sheikh Abdullah was against joining Pakistan. This
emboldened the government of the prime minister of India, Jawaharlal Nehru, on
December 31, 1947, to send a document to the UN Security Council asking it, in
essence, to play a key role in determining the sovereignty of Jammu and Kashmir
by holding and authoritatively supervising a plebiscite so that ‘‘its people would be
free to decide their future by the recognized democratic method of plebiscite or
referendum which, in order to ensure complete impartiality, might be held under
international auspices.’’ This offer was not totally unconditional. Nehru’s offer
stated that such a plebiscite should be held ‘‘after the soil of the State had been
cleared of the invader and normal conditions restored.’’38
The plebiscite was never held by the United Nations, or any other organiza-
tion, due to controversies and armed conflicts between Pakistan and India and
geopolitical complications relating to the Cold War.39 In this fundamental sense,
internationally disputed territory of Jammu and Kashmir was not a state in the way
we have defined it in the preface. Jammu and Kashmir was not like any other state
in the Indian federation in that not only was the adjudication of its sovereignty in
the hands of the United Nations but one third of its territory was also occupied by
a neighboring state, Pakistan. Jammu and Kashmir, in relation to the Indian state-
nation, from the beginning was therefore different and unique and critically did
not conform to the requirements of a state-nation.
To compound the problems of the geopolitical conflict, the government of
India again and again violated its own democratic procedures in the state of
Jammu and Kashmir. State-nation policies can work only in democratic condi-
tions. The absence of internationally recognized sovereignty constrained India
38. For three well-documented and somewhat different analyses of the international conditions
surrounding this conflict over Kashmir and Nehru’s invitation to the United Nations, see Gowher
Rizvi, ‘‘India, Pakistan, and the Kashmir Problem, 1947–1972,’’ Damodar R. Sardesai, ‘‘The Origins
of Kashmir’s International and Legal Status,’’ and Ashutosh Varshney, ‘‘Three Compromised Na-
tionalisms: Why Kashmir Has Been a Problem,’’ all in Perspectives on Kashmir: The Roots of the
Conflict in South Asia, ed. Raju G. C. Thomas (Boulder, CO: Westview Press, 1992), pp. 47–79, 80–
92, and 191–234, respectively.
39. In the judgment of one of India’s most experienced and distinguished diplomats a major
opportunity to hold the plebiscite and possibly end the sovereignty dispute that has contributed to
two more wars between India and Pakistan and killings and human rights violations by both
countries in Kashmir was lost by the government of India when they insisted that a plebiscite could
not be held unless India be allowed to have approximately ten thousand more Indian Army troops
on the ground in Kashmir during the plebiscite to compensate for armed groups supported by the
Government of Pakistan. Private conversation with Juan Linz and Alfred Stepan.
four indian cases that challenge state-nation theory? 111
from exercising the full repertoire of state-nation policies; the absence of demo-
cratic conditions in the state for most of post-Independence India did not make for
a context where the citizens could respond to the limited state-nation policies that
were tried out. It is of course a counterfactual, but if India had had uncon-
tested sovereignty, some combination of the solutions applied to the Punjab and
Mizoram could have been applied, enhancing the ‘‘we-feeling’’ with India and
the geopolitical security of the Indian citizens of Kashmir, Jammu, and Ladakh.
Article 370 of the constitution, discussed below, already provided the framework
for a state-nation solution. Undisputed sovereignty and democratic context could
have led to an effective assertion of the autonomy of the state, while safeguarding
the autonomy of the three regions within the state. The Dogri-speaking region
of Jammu in the southern highlands with a Hindu majority and the Tibetan-
speaking region in the former Buddhist kingdom of Ladakh bordering China in
the north could have been given special linguistic and cultural autonomy within
Jammu and Kashmir, along the same lines as the tribal districts within the state of
Assam. If these three cultural regions of Jammu and Kashmir had felt that they
were all securely part of the state-nation of India and allowed to exercise their full
state-nation democratic and cultural prerogatives, the sixty-year tragedy that is the
‘‘Kashmir crisis’’ may well not have occurred.
Some state-nation principles were nonetheless applied, at least in theory. On
paper, the institutional framework for division of power in India’s constitution
appears appropriate for Jammu and Kashmir. Sensitive to the special political
history of the state and keen to retain the only Muslim-majority province in the
Indian union, the framers of India’s constitution came up with Article 370, a
unique provision that carried the spirit of asymmetrical federalism to its limits
within the Indian constitutional order. This article, innocuously titled ‘‘Tempo-
rary provisions with respect to the State of Jammu and Kashmir,’’ granted the state
a special status, different from any other state in the Indian union. It stipulated
that barring defense, communication, and foreign relations, the Indian parlia-
ment would not be able to make laws for this state on any subject without the
concurrence of the state government.40 Jammu and Kashmir was the only Indian
40. The operational part of this article on ‘‘Temporary provisions with respect to the State of
Jammu and Kashmir’’ reads as follows:
(1) Notwithstanding anything in this Constitution,
...
(b) the power of Parliament to make laws for the said State shall be limited to,
(i) those matters in the Union List and the Concurrent List which, in consultation with the
Govermnent of the State are declared by the President to correspond to matters specified in the
Instrument of Accession governing the accession of the State to the Dominion of India as the
112 crafting state-nations
state to be allowed its own separate constitution, citizenship, and titles of some
officials. This provision was sharply opposed by Hindu nationalists; the abolition
of Article 370 has remained on their political agenda ever since the constitu-
tion was promulgated. The only thing missing from the constitutional provisions
was an arrangement to guarantee autonomy for the Jammu and Ladakh regions
within the state.
The state-nation approach followed in India meant that there was no attempt
to foster one language and one core cultural identity; this was as much true in
Jammu and Kashmir as it was elsewhere in India. The Kashmiri language was
included as one of the official languages of the Indian union, and in 2003 the
Dogri language from the Jammu region was included as well. But the uncertain
international status of Kashmir’s sovereignty contributed to Nehru authorizing
the arrest of Abdullah in 1953 to prevent him from leading a possible secessionist
movement in Kashmir. With rare exceptions, Abdullah remained imprisoned
until 1968.41 Since 1953, for long periods of time, the requirements of relative
political autonomy that are crucial for state-nation politics were arguably more
violated by the center with respect to Jammu and Kashmir than they were for any
other state in the Indian union. For more than two decades thereafter, Jammu and
Kashmir was governed by puppets of the Congress government in Delhi, who
enjoyed little popular legitimacy within the state. During this period, the state
legislature ceded to the parliament most of the powers it enjoyed under Article
370. No less than forty-eight presidential orders were issued, with the concurrence
of the state government, applying more and more provisions of the constitution to
Jammu and Kashmir. The state still has its own citizenship, a requirement for
holding land and other property, and its own constitution, but the special status
conferred by the constitution, for all practical purposes, no longer exists because it
has been so eroded by the central government.
Far from being a state with theoretically exceptional prerogatives and greater
autonomy than the rest, Jammu and Kashmir (largely due to its unique status
within the Indian federation as the only state subject to international dispute
about its sovereignty) actually became a specially disempowered state that was
practically run by the central government. The period between 1953 and 1974,
between Sheikh Abdullah’s arrest and his final release after a political settlement
matters with respect to which the Dominion Legislature may make laws far that State; and
(ii) such other matters in the said Lists as, with the concurrence of the Government of the State,
the President may by order specify.
41. For an excellent account of the deterioration of the relations between Nehru and Abdullah,
see Varshney, ‘‘Three Compromised Nationalisms,’’ pp. 191–224.
four indian cases that challenge state-nation theory? 113
with Indira Gandhi, witnessed a puppet government in Jammu and Kashmir and
proxy government by the Congress Party and the central government character-
ized by pliable chief ministers, state governors functioning as agents of the center,
restrictions on the operations of the National Conference, and attempts to break
the party. After a partial and short-lived restoration of autonomy following an
agreement between Sheikh Abdullah and Prime Minister Indira Gandhi in 1974,
the old order was restored in 1983. The National Conference was left with no
option but to ally with the Congress in the infamous 1987 election that was widely
perceived to have been rigged against the candidates of the Muslim United Front,
a coalition of pro-autonomy political forces. This was followed by a rise in mili-
tancy and an escalation in violence; the government of India then clamped down
and suspended normal political activities and imposed ‘‘President’s Rule,’’ which
means direct rule by the center.
This gross violation of political autonomy could take place only by denial of
basic democratic norms and freedoms within the state. It is widely accepted that
all of the elections that took place in the state between 1962 and 2002 were rigged,
with the exception of the assembly elections held in 1977 and, to some extent,
1983. Electoral fraud in Kashmir took many forms: banning parties and organiza-
tions that were not ‘‘acceptable,’’ not allowing opposition candidates to file nomi-
nations, rejecting their nominations on frivolous grounds, not allowing the op-
position to campaign freely, intimidating voters, forcing nonvoters to the polling
booths, encouraging multiple voting and mobile voters, stuffing ballots, and falsi-
fying vote counts.42 Since 1990 there has been an intermittent insurgency, making
the Indian Army, backed up by the notorious Armed Forces Special Powers Act
promulgated in the area that year, the most salient, and disliked, presence of the
government of India in Kashmir.
In a 2009 CSDS survey in Mizoram and the Punjab (after both had been
inclusive and peaceful state-nation polities for at least a decade), only 2% of
respondents in Mizoram and only 15% in the Punjab said they had ‘‘no trust’’ in
the central government. In sharp contrast, nearly two-thirds of the 1,118 people
polled in the Kashmir Valley in 2002 said that they had ‘‘no trust’’ in the central
government (88% said they had ‘‘no trust’’ in the Indian Army). In the same
survey, virtually no one in the Kashmir Valley expressed a preference ‘‘to stay with
India on present terms,’’ only 8% expressed a preference for becoming ‘‘part of
42. For an account of how the 2002 elections in the state were an improvement over the past, see
the book authored by former chief election commissioner James Lyngdoh, The Chronicle of an
Impossible Election: The Election Commission and the 2002 Jammu and Kashmir Assembly Elections
(Delhi: Viking India, 2004).
114 crafting state-nations
table 3.2
Persistent Alienation in the Kashmir Valley: Public Opinion on Solutions
to the Kashmir Problem in the Kashmir Valley, 2002–2008
In the present situation, which of the following is the most suitable option
for Kashmir? 2002 2008
‘‘Kashmir on both sides of the border should be merged and become an 79% 73%
independent country.’’
‘‘Kashmir should remain with India with maximum autonomy.’’ 6 16
‘‘Kashmir should remain with India as it is.’’ 1 1
‘‘Kashmir should become a part of Pakistan.’’ 8 2
No opinion 6 8
Source: Jammu and Kashmir Assembly Election Study, 2002 and 2008, post-poll surveys (sub-
sample from Kashmir Valley), CSDS, Delhi with Dept. of Political Science, Univ. of Jammu.
Note: That Jammu and Kashmir as a whole is not a state-nation polity with one overlapping
and complementary identity is clear from the responses in the Jammu and Ladakh regions to
the same question. In the 2002 survey only 7% of the respondents in the Jammu region and only
2% of the respondents in Ladakh supported the demand for merger of the two Kashmirs and
none supported merger with Pakistan.
Pakistan,’’ but a very strong 79% expressed a preference for Kashmir to become an
‘‘independent country’’ (see table 3.2).
The survey data in the Kashmir Valley in 2002 clearly indicate a picture of
political alienation, absence of identification with India, and very low trust in
some of the key institutions of India. We see this as a failure of international and
transnational political processes and of many Indian politicians. However, we do
not see this as evidence of the failure of state-nation politics, for two reasons.
First, the ‘‘scope conditions’’ for state-nation politics must be within the terri-
tory of an internationally accepted state. Only such a state has the authority
to help create and legitimate state-nation structures, practices, and principles.
Jammu and Kashmir has never been such a state, so from our theoretical perspec-
tive, it is outside the effective scope of the state-nation model.
Second, state-nation policies of the kind advocated here require a democratic
political setting. Largely because of its unresolved international ‘‘stateness’’ prob-
lem, Jammu and Kashmir has not enjoyed democratic institutions and practices.
The minimum conditions of relatively free and fair elections, civil and political
rights, and respect for democratically elected state government, which can be
taken for granted in the rest of the country, did not obtain in Jammu and Kashmir.
Although it was supposed to be the most autonomous part of the Indian federa-
tion, as originally laid out in the constitution, Jammu and Kashmir has de facto
been one of the least politically autonomous and one of the most centrally con-
four indian cases that challenge state-nation theory? 115
43. The case of Jammu and Kashmir also provides us with a temporal variation that can help us
test our argument. The political context changed significantly in the state after the assembly elec-
tion in 2002. This election was widely perceived to be the most free and fair election in the state
since 1977. Since then two parliamentary elections and one assembly election have been held in the
state, and the practice of largely free and fair elections appears to have been routinized. This has led
to a subtle change in political alienation and a separatist mindset within the Kashmir Valley. The
National Election Study 2004 records some of these changes. While nearly half of the sample of 954
respondents kept quiet on sensitive questions dealing with the fairness of the electoral process, an
overwhelming majority of those who gave a response acknowledged marked improvement: 76% said
that the 2004 parliamentary election saw virtually no rigging, 71% thought the level of electoral
malpractice had gone down, and 90% felt either no fear or said the fear had decreased. A repeat post-
poll survey in the state in 2008 recorded a sharp decline from 33% in 2002 to 22% in 2008 in
respondents from the valley who suspected rigging in the state assembly elections. More impor-
tantly, those respondents in the Kashmir Valley who favored the valley’s merger into Pakistan went
down from 8% in 2002 to just 2% in 2008, while those who were for Kashmir staying with India but
with greater autonomy went up from 6% to 16% during this period. However, the fragility of this
achievement has been underscored by massive and popular anti-India protests in the valley in 2010.
Between June and early October, more than one hundred civilians had been killed.
chapter four
Tamils in India
How State-Nation Policies Helped Construct
Multiple but Complementary Identities
There is an extremely long tradition in democratic social analysis that more or less
argues that the term multinational democracy is an oxymoron.1 In chapters 2 and 3
we presented compelling evidence to challenge that tradition, even in one of the
world’s most culturally diverse polities. Until now, except for our discussion of
some of the most distinctive features of the Indian polity, we did not go very far in
undertaking a theoretical and empirical inquiry into what helps increase or de-
crease the chances of democracy and social peace in polities with some multina-
tional dimension to their political life. This is our primary task in this chapter.
We will attempt to carry out this task by analyzing how a potential problem of
politically robust multinationalism with possible secessionist potential in South
India, the Dravidian movement, especially in what is now the state of Tamil Nadu,
became a non-problem.2
By Benedict Anderson’s standards, there would appear to have been more than
enough raw material for secessionist nationalists to ‘‘imagine a community’’ of
1. For a critical analysis of this political and intellectual tradition, see Stepan, ‘‘Modern Multi-
national Democracies: Transcending a Gellnerian Oxymoron,’’ in his Arguing Comparative Politics
(New York: Oxford University Press, 2001), pp. 181–199.
2. For one of the most cited books about the Tamil secession as a potential problem, see Eu-
gene F. Irschick, Politics and Social Conflict in South India: The Non-Brahmin Movement and Tamil
Separatism, 1916–1929 (Berkeley: University of California Press, 1969). For two important reviews of
the literature of the Dravidian movements, see M. S. S. Pandian, ‘‘Beyond Colonial Crumbs:
Cambridge School, Identity Politics, and Dravidian Movement(s),’’ Economic and Political Weekly,
February 18–25, 1995, pp. 385–391; and N. Ram, ‘‘Dravidian Movement in its Pre-Independence
Phases,’’ Economic and Political Weekly, vol. 14, no. 7/8 (February 1979), pp. 377–397.
tamils in india 117
their own that could be a separate independent nation in what is now South
India.3 Useable cleavages abounded.
A potentially useable cleavage grew out of religious-cultural differences. In the
south, the Brahmins were seen as northern in origin. Nationalists in the south,
particularly near the important city of Madras (now Chennai), argued that tradi-
tional Dravidian culture had been more socially egalitarian than the version of
Hinduism imported and imposed upon Dravidians by northern Brahmins. The
potential of caste to be a polarizing force was enhanced by the fact that under
British rule Brahmins were accorded a new higher social status that in effect
lowered the social status of some South Indian caste groups that had been socio-
economically and religiously important.4 For some analysts, since the two inter-
mediate Hindu castes, Kshatriya and Vaishya, were virtually not present in South
India, the South Indians therefore either belonged to the lowest category of caste
Hindu—the Shudras—or were untouchables. This could have increased the so-
cial and political distance of southern Indians from northern ones.5
Modernity, à la Gellner, sharpened South Indians’ sense of exclusion and
contributed to the expanding anti-Brahmin movements. The emerging Dravidian
nationalist movements in the early decades of the twentieth century gained ad-
herents as they documented and dramatized job-related statistics aiming to prove
that non-Brahmins were second-class citizens in South India. For example, the
famous ‘‘Non-Brahmin Manifesto of 1916’’ argued that although Brahmins con-
stituted less than 3% of the population in the major administrative sub-unit of
South India, the Presidency (state) of Madras, all but one of the sixteen top civil
service positions allocated to Indians in Madras were held by Brahmins, all four of
the Hindu judges to the Madras Supreme Court were Brahmins, and the major
gate-keeper of modern careers, the University of Madras, was effectively con-
trolled by Brahmins.6
Language was a compounding factor that differentiated southern Indians from
northern Indians. In the last decades of the British Raj, more than 90% of the
3. Benedict Anderson, Imagined Communities: Reflections on the Origin and Spread of Na-
tionalism (London: Verso, 1983).
4. See Nicholas B. Dirks, Castes of Mind: Colonialism and the Making of Modern India (Prince-
ton: Princeton University Press, 2001), esp. chs. 1 and 12.
5. For this argument, see Marguerite Ross Barnett, The Politics of Cultural Nationalism in South
India (Princeton: Princeton University Press, 1976), pp. 46–47.
6. The manifesto is reproduced in its entirety in Irschick, Politics and Social Conflict in South
India: The Non-Brahmin Movement and Tamil Separatism, 1916–1929, pp. 358–367. From 1901 to
1911, Brahmins received 71% of the degrees awarded by Madras University and controlled the key
power center in the university, the Senate. See Barnett, Politics of Cultural Nationalism, p. 20.
118 crafting state-nations
population in South India spoke languages in the Dravidian family, each with a
distinct script and unintelligible to users of the major language of the north,
Hindi.7 At Independence, the geographic, demographic, and imagined space of
South India contained 88 million speakers of Dravidian languages. The four
largest of these Dravidian languages in 1951 were Telugu (33 million), Tamil (27
million), Kannada (14 million), and Malayalam (13 million). The cultural capital
of the Tamils was the city of Madras in what was then called the Madras Presi-
dency, but the Dravidian movement also contained important advocates from the
other three major Dravidian languages, some of whom also lived in the Madras
Presidency.8
Would be secessionist nationalists had other valuable material. They could
point out that they were economically more developed than the Hindi Belt of
northern India, which they saw as politically dominant; that South India was
geographically contained, in that two of its three borders were seas; and that it was
populous enough to make one or more South Indian independent nations.
A leading scholar of India, Lloyd I. Rudolph, did extensive research in South
India in the 1950s, graphically capturing how the different components of seces-
sionist nationalism seemed to be accumulating at the time of Independence:
‘‘With the coming of independence, anti-Brahmanism was increasingly accom-
panied by an anti-North, Dravidian nationalist outlook. Opposition to Hindi as
the national language, the destruction of the caste system, and threats of secession
from the Indian Union became major political themes.’’9
In politics, facilitating structures normally do not become actualized realities
without agents. There were powerful agents. One such agent was the charismatic,
autocratic, Dravida nationalist leader Ramaswami Naicker (usually called ‘‘Periar’’
or ‘‘Periyar’’), whom a leading specialist called ‘‘one of the most dynamic and
colourful political leaders South India has ever produced.’’10 A leading scholar of
twentieth-century Tamil Nadu, Narendra Subramanian, asserts: ‘‘Periar called for
the creation of a separate country in which the Dravidian-as-Sudra would enjoy
primacy.’’11 Another important agent was C. N. Annadurai, who broke with Periar
7. Jyotyrindra Das Gupta, Language Conflict and National Development: Group Politics and
National Language Policy in India (University of California, Berkeley. Center for South and South-
east Asia Studies, 1970), pp. 46–47.
8. For the list of the major languages of India as of 1951, see ibid.
9. Lloyd I. Rudolph, ‘‘Urban Life and Populist Radicalism: Dravidian Politics in Madras,’’
Journal of Asian Studies 20, no. 3 (May 1961), pp. 286–287.
10. See Robert L. Hardgrave Jr., ‘‘Religion, Politics, and the DMK’’ in South Asian Politics and
Religion, ed. Donald Eugene Smith (Princeton: Princeton University Press, 1966), pp. 216, 223.
11. Narendra Subramanian, Ethnicity and Populist Mobilization: Political Parties, Citizens, and
Democracy in South India (Oxford: Oxford University Press, 1999), p. 105.
tamils in india 119
Rikerian Federalism,’’ in his Arguing Comparative Politics, and his ‘‘Federalism and Democracy:
Beyond the U.S. Model,’’ Journal of Democracy 10, no. 4 (1999), pp. 19–34.
17. Ambedkar, Address to the Constituent Assembly. Ambedkar’s speech is found in its entirety
in India, Constituent Assembly Debates (New Delhi, 1951), vol. 2, pp. 31–44, emphasis added.
18. Mohit Bhattacharya, ‘‘The Mind of the Founding Fathers,’’ in Federalism in India: Origins
and Development, ed. Nirmal Mukarji and Balveer Arora (New Delhi: Vikas, 1992), pp. 87–104,
quotations from pp. 101–102. Bhattacharya’s language concerning the function of the type of federal-
ism adopted in India approaches our ideal type of ‘‘holding together’’ federalism.
tamils in india 121
(2) uniformity in fundamental laws, civil and criminal, and (3) a common All-
India Civil Service to man important posts.’’19
In relation to the demos-constraining versus demos-enabling continuum, India
chose one of the most demos-enabling formulas found in any democratic federa-
tion, be it mononational or multinational. The U.S. formula, which did grow out
of a ‘‘coming together’’ bargaining process, gave each state equal representation in
the upper house and gave the upper house somewhat greater legal competencies
than the lower house. India’s ‘‘holding together’’ federation was fundamentally
different in both respects. The lower chamber, which was based on the principle
of population, had the exclusive right to form the government and thus was vastly
more important in legislative competence than the upper chamber, which repre-
sented the states. Furthermore, there was a significant degree of proportional
representation in the upper chamber. The demos at the center, aided by the
choice of a Westminster type of fused executive-legislative parliamentary model,
was thus not nearly as constrained in independent India as was the demos at the
center in the presidential and divided-government model chosen in the United
States. In fact, as Stepan has documented, the United States has four electorally
generated ‘‘veto players’’ in that the Senate, the House of Representatives, the
president, and, in the matter of amendments, the states can all exercise powerful
vetos. In contrast, India is like Austria in that the only veto player is the lower
house of parliament.20
A major controversy in the Constituent Assembly was over the languages that
would be used in the federation. Precisely because the members of the Con-
stituent Assembly knew that the most controversial issue surrounding Indian unity
in the future would be language policy and because there was a desire on the part
of many delegates to reorganize the states at some point along linguistic lines,
the language of the constitution was extremely demos-enabling for the lower
house of the federal parliament, the Lok Sabha. Future parliaments were given
the right to redraw state boundaries as they wished. Article 3 of the constitution is
categorical on this point. With a simple majority, ‘‘Parliament may by law a) form
a state by separation of territory from any state or by uniting two or more states . . . ;
. . . c) diminish the area of any state . . . ; . . . e) alter the name of any state.’’ In
a ‘‘coming together’’ federation such as the United States, the sovereign states
would obviously have been able to bargain successfully for a much more demos-
constraining constitution to protect states’ rights.21
That the demos, as represented in the Constituent Assembly, gave the parlia-
ment the right to work with the numerous linguistic demoi of India to restructure
the states turned out to play an important role in allowing the demos of India, and
the demoi of India, to ‘‘hold together’’ in a multinational democratic federal sys-
tem. We will not retell that story because it has been excellently analyzed by other
writers.22 The key point to stress here is that in 1955 a State Reorganization Com-
mission was formed to suggest changes in state boundaries. As a result of that
commission, eventually most of the units of the Indian federation were geograph-
ically and sociologically reconfigured to achieve greater congruence between lan-
guages and state boundaries. Each state was allowed to carry out its state admin-
istration in the dominant language of the state. This major constitutional change
meant that a significant degree of politically legitimated linguistic and cultural
nationalism had been achieved and recognized inside India’s federal polity.23
Das Gupta’s classic book on Indian national language policy concludes with
an analysis of the integrative effects of India’s recognition of the Tamil language as
the language of self-government in the federal unit of Tamil Nadu:
In Madras, most of the leaders of the DMK were once associated with the
demand for a homeland for the Dravidians. But as soon as they discovered that
they too were capable of winning the elections and capturing the political
authority of the state, they gave up their secessionist associations. . . . [L]anguage
politics has given a new meaning to the political community in India. It has
indicated that a viable political community can be built in India on the basis of
the recognition of the separate but related language communities. . . . [L]an-
guage politics has proved to be one of the most important positive democratic
channels for pursuing political integration.24
21. See Stepan, ‘‘Federalism and Democracy: Beyond the U.S. Model,’’ Indeed, it was precisely
this feature of the Indian constitution that then led the leading theorist of federalism in the world, K.
C. Wheare, to argue, ‘‘What makes one doubt that the Constitution of India is strictly and fully
federal, however, are the powers of intervention in the affairs of the states given by the Constitution
to the central government and parliament. To begin with, the parliament of India may form new
states; it may increase or diminish the area of any state and it may alter the boundaries or name of
any state.’’ See K. C. Wheare, Federal Government (Oxford: Oxford University Press, 1963), p. 27.
22. See Das Gupta, Language Conflict and National Development, p. 33; and Paul Brass,
Language, Religion, and Politics in North India (Cambridge: Cambridge University Press, 1974).
23. Any such restructuring of course creates some new linguistic minorities. In contrast to Spain
and Belgium, the central government in India is mandated with the task of ensuring that these
linguistic minorities have sufficient schools in their own language.
24. Das Gupta, Language Conflict and National Development, pp. 268, 269, and 270. Unlike
tamils in india 123
Clearly, the language policies that Das Gupta is describing here are state-
nation policies nested inside India’s ‘‘holding together’’ federalism. Let us now
turn to another part of our nested grammar, one that concerns the role of polity-
wide parties introducing multiple but complementary identities.
In the immediate pre- and post-Independence era, what was the relationship
between culturally nationalist ‘‘regional parties’’ or movements and a ‘‘polity-
wide’’ party? In 1944 an offshoot of previous Dravidian movements, the Self-
Respect Movement and the Justice Party, was renamed the Dravida Kazhagam
(DK). According to Irschick, the DK had ‘‘as its primary aim the realisation of a
separate non-Brahmin or Dravidian country.’’25
For our analytical purposes, it is important to note that this regional na-
tionalist movement had to compete with the polity-wide party, the Indian Na-
tional Congress. The Congress Party had an All-Indian institutional presence
since its formation in 1885. This polity-wide party had acquired great legiti-
macy and experience owing to its leadership role in the Indian Independence
movement and the mobilizing capacity of Mahatma Gandhi. However, the Con-
gress Party originally made the mistake of recruiting most of its leaders in the
Madras Presidency from the small and culturally alien Brahmin community.26
Nonetheless, the regional nationalist movement led by Periar, the DK, never
became a political party. In 1949 the DK lost some important followers when an
equally charismatic but more democratic leader, C. N. Annadurai, left the DK to
form a political party called the Dravida Munnetra Kazhagam. The DMK was
not yet able to compete successfully with the Congress Party in the founding
polity-wide elections. In the first post-Independence elections in 1952, the Con-
gress Party won twelve of the fourteen seats for the parliament in Delhi from
Madras but only a plurality, 152 of 375 seats, to the Madras state assembly. The
DMK, even though it was a culturally nationalist party that still had not publicly
renounced all secessionist possibilities, supported a United Democratic Front
Spain, where well over 90% of the population shares a mutually comprehensible language, Spanish,
even if their first language might be Basque or Catalan, in India more than 50% of the population
speaks neither Hindi nor English.
25. Irschick, Politics and Social Conflict, p. 347.
26. Indeed, the previously cited ‘‘Non-Brahmin Manifesto’’ explicitly laments that fourteen of
the fifteen members of the Madras Congress Party Committee were Brahmins. See Irschick, Politics
and Social Conflict, p. 361.
124 crafting state-nations
coalition of parties, some of which, the Communists and the Socialists, were
polity-wide parties.
The original leader of the Congress-led government in Madras after the first
post-Independence elections of 1952 was a Brahmin, C. Rajagopalachari, who was
perceived to be insensitive to lower-caste and Tamil cultural aspirations. But in
the new electoral context, which had become very competitive, Kamaraj Nadar, a
lower-caste, Tamil-speaking, professional Congress Party organizer, became a cru-
cial leader linking Tamils and the Congress Party. Kamaraj did not have the
benefit of much formal education and did not speak Hindi or English, but he
combined strong Indian nationalist and Tamil cultural-nationalist roots, making
him a classic example of a state-nation political leader with multiple but comple-
mentary identities. Kamaraj, who was imprisoned six times by the British and had
spent more than three thousand days in jail for his pro-Independence activities,
emerged as the kingmaker in the Madras Congress Party. By 1954 he had become
the chief minister of Madras. Significantly, there was not one Brahmin in his first
cabinet.27
Jawaharlal Nehru employed a leadership style that, both as prime minister of
the government and as president of the Congress Party, relied heavily on the
consensual support of regional leaders. Nehru and Kamaraj related to each other
in ways that prevented center-periphery relations from being a zero-sum game. As
a major party regional boss, Kamaraj had political resources at the center. Indeed,
Kamaraj became one of the five members of the group called ‘‘the syndicate’’ that
co-ruled the Congress Party with Nehru. Moreover, after the death of Nehru,
Kamaraj became the president of the Congress Party. Kamaraj was effective as a
leader of a polity-wide party partly because his autonomy as a Tamil political and
cultural leader won the respect of Nehru. Power for both leaders was thus non-
zero-sum. Nehru, by ideological preference, would have preferred a strong Indian
central government that generated an increasingly homogeneous nation-state
culture. Politically, however, he knew that he had to depend on a core of Congress
Party members who represented, and led, India’s major regions with their diverse
languages and cultures. In essence, Nehru followed a ‘‘strong centre, strong sub-
unit’’ policy.
In this type of federal politics, Kamaraj was a regional leader in Tamil-speaking
India who had sufficient strength and respect in the center for him to be allowed
to deliver many cultural-nationalist demands. But as a regional boss of a large
27. See Duncan B. Forrester, ‘‘Kamaraj: A Study in Percolation of Style,’’ Modern Asian Studies
4, no. 1 (1970), p. 49.
tamils in india 125
state, Kamaraj could also deliver valuable votes and support to the polity-wide
party. Lloyd Rudolph summarizes Kamaraj’s contribution to the strength, inside
the state of Madras, of the polity-wide Congress Party: ‘‘Between 1952 and 1957
Congress increased its share of the popular vote from 35.5 percent to 45.3 percent
largely by identifying itself more closely with the [Tamil] populist appeal. . . . The
growth in Congress strength can be attributed largely to the leadership qualities of
Mr. Kamaraj.’’28
In our discussion of the new ‘‘grammar’’ of federalism, we argued that it is
necessary to analyze some important democratic federations such as Spain, Bel-
gium, Canada, and India in their multinational context. Thus we stressed the
importance of ‘‘multiple but complementary identities.’’ We think this concept
is valid, and indeed necessary, but perhaps our phrase does not quite capture the
dual but nonetheless occasionally competing identities many nationalists might
feel. Most Tamil-speakers were very interested in Tamil cultural-nationalist goals;
from the 1930s on, many were members of parties or movements that periodically
articulated separatist aspirations. However, many Tamils were also interested in
the struggle for Indian independence. Since the most effective mass-based, pro-
independence organization was the polity-wide Congress Party, many Tamil cul-
tural nationalists identified with the Congress Party and never became secession-
ist nationalists.29 As the chief minister of Madras, Kamaraj (and the Congress
Party) received some cultural-nationalist credit for the creation of special quotas
for lower-caste Tamils and for their support for the Tamil language. A leader like
Kamaraj reduced the potential tension between polity-wide and the cultural-
nationalist goals. But without a polity-wide party, he could not have played such
a role.
That the three other major Dravidian-speaking areas of India had, by the
late 1950s, also been given a state in which the language of the government
was their own ended any possibility of a successful movement for an indepen-
dent Dravidian country. This was so because by the late 1950s linguistic cultural-
nationalist claims for the then 37 million Telegu speakers were organized and
28. Rudolph, ‘‘Urban Life and Populist Radicalism,’’ p. 294. On the ‘‘cultural nationalist’’ and
‘‘polity-wide party’’ appeals of Kamaraj, see also Hardgrave, ‘‘Religion, Politics, and the DMK,’’ pp.
226–227.
29. In electoral terms, in the 1920s the Congress Party, with its pan-Indian ideology, and the
Dravidian cultural nationalist Justice Party would have seemed to be in a zero-sum relationship.
However, the existential reality of people who simultaneously wanted to affirm support for cultural
nationalism and pan-Indianism is beautifully shown by a quotation from Subramanian: ‘‘Congress
was so popular that by 1927 the Justice Party was forced for reasons of survival to allow its members to
have parallel membership in Congress.’’ Ethnicity and Populist Mobilization, p. 125.
126 crafting state-nations
articulated by the state of Andhra Pradesh; for the 30 million speakers of Tamil by
the state of Madras (which changed its name to Tamil Nadu in 1968); for the 17
million Kannada speakers by the state of Karnataka; and for the 17 million Ma-
layalam speakers by the state of Kerala.30
But, if the States Reorganisation Commission Report effectively ended any
chance of a united, separatist, Dravidian movement for a single country, there
were still some advocates of a territorially independent country of Tamil Nadu.
Why, then, did the Tamil cultural nationalists defeat the Tamil secessionist na-
tionalists?31
The constitutional decision to make India a federation made it possible for politi-
cal activists like Kamaraj to be a cultural-nationalist leader at the state level and an
All India leader at the center. Furthermore, the decisions to craft a ‘‘demos en-
abling’’ and ‘‘asymmetrical’’ federation allowed the parliament at the center to go
forward with the fundamental redrawing of the political boundaries of the federa-
tion to reflect the cultural-nationalist demands of language. And, of course, the
political activity and organization of the Congress Party since 1885 allowed a
polity-wide party to compete effectively in the elections against regional-cultural
nationalists, even in the newly created linguistic states.
The creation of a Tamil-speaking state in a context where cultural national-
ism was very strong gave the two Tamil nationalist political organizations, the
DMK and the DK, an opportunity to win control of the state by waging cultural-
nationalist campaigns. The DMK participated in the election for state and federal
legislatures in 1957, but due to the popularity of Congress and Kamaraj, it did not
do very well. The DK, which was not a party, did not compete in the election but
continued with its formal demands for a sovereign and independent country.
In 1959, the DMK, with an adroit campaign focused on local government,
finally won political control of Madras, the largest city and capital of the state of
Madras.32 After 1959 the Tamil nationalist DMK increasingly began to believe it
could win control of the state assembly and the state government, and some of its
30. The number of speakers of these languages is from the Census of India, 1961, reproduced in
Das Gupta, Language Conflict and National Development, p. 46.
31. Juan J. Linz, ‘‘From Primordialism to Nationalism,’’ in New Nationalisms of the Developed
West, ed. Edward A. Tiryakian and Ronald Rogowski (Boston: Allen and Unwin, 1985), pp. 203–253.
32. Barnett, Politics of Cultural Nationalism, p. 105.
tamils in india 127
leaders and followers even harbored ambitions for greater political autonomy.
However, DMK parliamentarians in Delhi sent back warnings that separatist
parties might be made illegal and that secessionist nationalist demands in the state
of Madras were unfeasible and dangerous. The ‘‘no exit’’ dimension of Indian
statehood we discussed previously was supported by the Indian Army, which,
given its loyalty to the center and great size, was a completely credible coercive
force.33 Thus, not wanting to jeopardize their chances to win control of the state of
Madras, top DMK leaders in 1960, in a closed private meeting, made a decision to
drop their secessionist nationalist aspirations, implicitly but not explicitly.34
In the general election of 1962, the DMK emerged as the major opposition. In
the 1962 campaign the DMK ran as a cultural-nationalist party that had not
formally abandoned secessionist nationalism, but secessionist demands did not in
fact figure prominently in its campaign. In 1967, the DMK defeated the Congress
Party and won control of the state of Madras.35 From 1967 on, the DMK never
gave up its cultural nationalism, but it did become increasingly integrated into the
politics and norms of the Indian federation.
One of the elements that can help sustain complementary identities in a polity
with some multinational dimensions is the role of material interests, especially
accommodative language policies that enhance polity-wide career opportunities.
In 1965 and 1970, there were two protest movements in Tamil-speaking India.
Both were, in essence, struggles to maintain India-wide career opportunities. In
one case there was a struggle against an expected change in official language
policy of the central government for the whole country. In the other case, the
struggle was against the excessive culturally nationalist language policies of the
regional government. In both cases, some of the same activists participated.
The first protest movement was a response to the fifteenth anniversary of the
constitution of India, which took place on January 26, 1965. The 1950 constitution
33. The ‘‘no exit’’ situation was made viable due to the absence of external state support for any
secessionist movements and the probability that if the central government used force, it would not
be subject to any significant international censure or sanctions.
34. Ibid., pp. 102–115.
35. For an account of the growing integration of Tamil politics into Indian federal politics by a
leading specialist on Indian federalism, see Balveer Arora, Specificite Ethnique, Conscience Region-
ale et Developpment National: Langues et Federalism en Inde (Thèse pour le Doctorat de Re-
cherches, Fondation Nationale des Sciences Politiques, Paris, 1972), esp. pp. 193–406. See also
Subramanian, Ethnicity and Populist Mobilization, pp. 160–172.
128 crafting state-nations
stipulated that ‘‘it was the duty of the Union to promote the spread of the Hindi
language, to develop it so that it may serve as a medium of expression for all
elements of the composite culture of India.’’36 Moreover, English was accepted as
a useable ‘‘link language’’ of the federation for fifteen years, which were up in
1965. There were widespread demands in the northern Hindi heartland, even
among the Socialists, to make Hindi the official language of the union. In the
south, particularly among Tamil elites, there were intense fears that Hindi would
become the only acceptable language for entrance examinations to the coveted
and powerful Indian Administrative Service and for India’s court system. A long-
time observer of South Indian politics based in Madras in 1965 nicely captured
elite and middle-class fears about career prospects in this period: ‘‘Students, law-
yers, and businessmen, indeed the Madras middle class generally, see their inter-
ests as tied to the continuance of English as the medium for the Union Public
Service Commission’s competitive examinations. Northerners and Southerners
start from the same point in English; the introduction of Hindi would impose a
serious hardship on those for whom it is not their mother tongue.’’37
Faced with the threat of losing these polity-wide career opportunities, students,
supported by lawyers and many other groups, waged for much of January and
February 1965 the biggest protests in Madras since the anti-British ‘‘Quit India’’
protests of the 1940s. These protests rapidly became riots, and police and army
troops opened fire in twenty-one towns in the state, arrested over ten thousand
people, and reportedly killed over one hundred people.38 The two Tamil-speaking
ministers of the central government (for agriculture and petroleum) submitted
their resignations.
To stop this growing crisis of the Indian state-nation, on February 11, 1965,
Prime Minister Shastri announced a crucial decision on an India-wide broadcast:
‘‘For an indefinite period . . . I would have English an associate language . . . I do
not wish the people of the non-Hindi areas to feel that certain doors of advance-
ment are closed to them. . . . I would have [English] as an alternative language as
long as people require it, and the decision [to maintain or revoke English as a link
language] I would leave not to the Hindi-knowing people, but to the non-Hindi
knowing people.’’39 More than forty-five years later, Shastri’s decision remains the
de facto policy of the federal government. Both the protests and the center’s
reaction contributed to the maintenance of polity-wide careers that help under-
gird multiple but complementary identities in Tamil Nadu.
What if Hindi had been imposed as the only official language of the Indian
federation in 1965? Stepan asked this question to C. Subramanian, one of the
Tamil-speaking ministers who had submitted his resignation. Subramanian re-
sponded that the president of India virtually refused to accept his resignation and
asked Shastri, ‘‘Do you want to lose Tamil Nadu from India? If not, kindly take
back your recommendation.’’40 Subramanian went on to speculate that if Hindi
had been imposed and English had been eliminated as a link language, the
protest movements would have been more virulent and the dormant secessionist
movement would have suddenly become greatly reinvigorated and possibly have
won.41 It is impossible to say if C. Subramanian’s speculations would have been
borne out. However, the DMK, which, as we have seen, had become cultural-
nationalist, might have faced increasing pressure to reintroduce a secessionist
discourse so as not to lose control over Tamil nationalism. Significantly, Barnett
notes that during the anti-Hindi mobilizations, the DMK, for the first time in
many years, actually did lose control over the leadership of the most important
Tamil and Dravidian protest movements and could not keep them within consti-
tutional limits.42 At the very least it would appear that the combination of Hindi
imposition and the removal of English as a useable language for civil service
examinations would have been a major blow, as in Sri Lanka, to the possibility of
polity-wide careers and multiple but complementary identities that are so useful
in maintaining peaceful and democratic federalism in multinational settings.
In 1967, the DMK won the provincial elections and became the first cultural-
nationalist party to assume control of an Indian state. The question of career
opportunities once again assumed great importance, but in this case students put
pressure on the DMK’s chief minister, C. N. Annadurai, not to close off their
career paths in the Indian polity and market. Here the desire of a political party’s
39. Cited in Barnett, Politics of Cultural Nationalism, p. 134. On the 1965 protests, see pp. 131–
135 in Barnett’s book as well as Richard L. Hardgrave Jr., ‘‘The Riots in Tamilnad: Problems and
Prospects of India’s Language Crisis,’’ Asia Survey 5 (August 1965), pp. 399–407; and Forrester,
‘‘Madras Anti-Hindi Agitation, 1965.’’
40. Interview of Alfred Stepan with C. Subramanian, Chennai, Madras, April 1, 1998.
41. Ibid.
42. Barnett, Politics of Cultural Nationalism, pp. 132–135.
130 crafting state-nations
Prime Minister Jawaharlal Nehru died in 1964 and closely fought elections be-
came increasingly important after 1967. Indeed, the Congress Party lost power in
what is now Tamil Nadu in 1967 and has never again formed the government by
itself in that state. However, in the vast majority of states, India’s combination of
numerous political parties and a first-past-the-post, single-member-constituency
electoral system means that since 1967 a single party running alone often loses to a
candidate supported by a multi-party alliance, and a single party by itself normally
does not get a majority in the provincial legislature, which would allow it to form a
single-party government.
As long as the above conditions exist, nested in a parliamentary context, and as
43. Ibid., p. 291. For the tone of this conflict, which was shorter and less intense than that of 1965,
see two front-page articles in The Hindu, January 1, 1971. In Spain, the policy effort of the Catalan
government to impose a predominantly Catalan education system up to the university level does not
generate an opposition of the large Spanish-speaking minority in metropolitan Barcelona because
its members believe that the Spanish language is so present in their daily lives that it is not threat-
ened as a language.
44. Ibid., p. 291.
tamils in india 131
long as state and federal elections are held and offices and appointments flow
from electoral results, there will be strong incentives to form multi-party electoral
coalitions.45
Surprisingly, this proposition holds even for coalitions that combine polity-
wide parties and parties that would seem to be potentially separatist regional
parties. But, and this is the crucial point, the mutual electoral benefit of coalitions
can only be obtained if both the potentially separatist regional parties and the
polity-wide parties adjust their behavior (and votes) to make the alliance possible.
The incentive system of this type of electoral bargaining works like this: A
polity-wide party would be severely constrained against entering into an electoral
alliance with a secessionist nationalist party, which articulates, or is widely be-
lieved to harbor, secessionist ambitions, because it would be attacked throughout
the rest of India by polity-wide parties for contributing to the ‘‘disintegration’’ of
India. For its part, a regional cultural-nationalist party would be severely con-
strained against entering into an alliance with any polity-wide party that voted in
the federal legislature for the imposition of assimilationist policies, because it
would fear losing votes to other culturally nationalist parties.
Tamil Nadu in 1971 illustrates the complex electoral and policy tradeoffs that
can make an apparently cultural-nationalist ‘‘regional’’ party, in effect, ‘‘centric-
regional,’’ and a ‘‘polity-wide’’ party, in effect, supportive of regional cultural
nationalism. Let us explore this complicated—but absolutely crucial—aspect of
Indian federalism.
By the late 1960s the Congress Party had split into a Congress (R) faction led by
Indira Gandhi and a Congress (O) faction. The Tamil cultural-nationalist parties
had also divided into the DMK and another group that later became the Anna
Dravida Munnetra Kazhagam (ADMK) in 1972. Congress (R) was primarily inter-
ested in how it did in the federal Lok Sabha elections, and the DMK was primarily
interested in how it did in the Tamil Nadu state assembly elections. But both the
Congress (R) and the DMK felt they would be greatly helped by forming a strong
government in their respective spheres of greatest interest if they could work out
45. See E. Sridharan, ‘‘The Fragmentation of the Indian Party System, 1952–1999: Seven Com-
peting Explanations,’’ in Parties and Politics in India, ed. Zoya Hasan (Oxford: Oxford University
Press, 2002). See also Balveer Arora, ‘‘Negotiating Differences: Federal Coalitions and National
Cohesion,’’ in Transforming India, ed. Francine Frankel, Zoya Hasan, Rajeev Bhargava, and Balveer
Arora (Oxford and Delhi: Oxford University Press, 2000), pp. 176–206; and K. K. Kailash, ‘‘Coali-
tions in a Parliamentary Federal System: Parties and Governments in India,’’ unpublished Ph.D.
dissertation, Jawaharlal Nehru University, 2003. The major exception to this logic of incentives to
form a coalition might be if a ‘‘regional secessionist’’ party controlling a government of a state was,
for whatever reason, financially and politically independent of the center and did not want to enter a
coalition at the center as a matter of principle.
132 crafting state-nations
Although the DMK alliance with Indira Gandhi’s Congress (R) seems paradoxi-
cal, given previous DMK separatist tendencies, it is in fact consistent with DMK
priorities and cultural nationalist orientation. In analyzing DMK political ac-
tivities and policies on the national level, it is essential to remember that the
primary party priority was consolidation of their state-level base.
After the November 1969 Congress party split, a unique opportunity was
46. The 1971 election was not an exception but a precedent. Since 1977 the DMK has had to
compete with a spinoff party, the ADMK, later called AIADMK. In the vast majority of the elections
since 1977, one or both of these once-Dravidian parties have been in alliances with non-Dravidian,
polity-wide parties.
tamils in india 133
created for the DMK to enhance its national image, improve relations with the
centre, and most importantly, consolidate its state support base by linking itself
to the left-leaning economic and social policies of Indira Gandhi.
Since 1971 the DMK has been solidly ‘‘centric-regional.’’ Given the coali-
tional incentive system we have just described, even the DMK’s major
cultural-nationalist competitor, the All India Anna Dravida Munnetra Kazhagam
(AIADMK), routinely enters into alliances with polity-wide parties, so it, too, is
subject to the same ‘‘centric-regional’’ incentive system.
47. Thus, Narendra Subramanian, writing about Tamil Nadu, correctly asserts that ‘‘the mate-
rial interests of many core DMK supporters were not directly linked to secessionism.’’ See his
Ethnicity and Populist Mobilization, p. 313.
48. Forrester, ‘‘Madras Anti-Hindi Agitation, 1965,’’ p. 34.
134 crafting state-nations
de facto permanent link language, and the Tamil language was secured as the
language of regional power, virtually no key Tamil leaders ever again spent major
resources on the goal of achieving independence and a separate independent
country. In fact, in terms of the definitions advanced earlier, Tamil India re-
mained deeply multicultural but was no longer ‘‘robustly multinational.’’
Barnett administered a poll in Tamil Nadu to DMK and Congress Party activ-
ists in 1968. Of the 459 local DMK party leaders who were asked the question
‘‘What do you consider the most important problem in your district?’’ only 2%
mentioned issues of language as the first problem, and none mentioned indepen-
dence. Of the thirty-eight members of the DMK General Council who were
asked about their reasons for being active in the DMK movement, none men-
tioned Tamil Nadu separatism as the first reason, but 7.8% did mention Tamil
language and culture, and 10.5% mentioned the two-language policy.49
In this context, worries about cultural nationalism or threats to integration
were not salient even for the 120 state-level Congress Party leaders interviewed in
Tamil Nadu. Indeed, in 1968 only 2.5% listed as their first worry threats to na-
tional integration.50 The responses to these questions, by both the DMK ‘‘centric-
regional parties’’ and the Congress ‘‘polity-wide’’ party activists, are further support
for our overall argument that the potential issue of separatism in Tamil Nadu had,
by the early 1970s, become a non-issue in India’s state-nation.
When this chapter appeared in 2003 as a freestanding paper, it ended with the
above discussion of the state of politics in Tamil Nadu in the late 1960s. However,
since then, we have all participated in preparing some questions for the Indian
National Election Study, 2004, and the five-country State of Democracy in South
Asia survey. Since the original text stands (virtually) as written in 2003, the incor-
poration of these new survey results thus represents a postscript. The surveys are
an invaluable source of information about opinions in Tamil Nadu toward India
forty years after the events described in this chapter. We should also add that the
three of us worked particularly hard to write questions that would shed light on
our conceptual approach to state-nation politics and other issues that we consider
central to our argument. Do these survey findings bolster or undermine our
table 4.1
Vote Share of Polity-wide Parties in Tamil Nadu State Assembly Elections, 1952–2001
Vote share of
Election year polity-wide parties
1952 63.3%
1957 55.3
1962 63.9
1967 50.1
1971 43.2
1977 55.6
1980 29.7
1984 24.4
1989 26.5
1991 23.4
1996 13.0
2001 9.1
Source: CSDS Data Unit, Delhi, based on the official returns of the Election Commission of
India.
Note: The table’s identification of polity-wide parties follows the Election Commission of
India’s official classification of ‘‘recognised national’’ parties. This did not apply in the first
general election of 1952. In that election, we have identified the following parties as ‘‘national’’:
Indian National Congress, Communist Party of India, Socialist Party, Kisan Majdoor Praja
Party, Bhartiya Jana Sangh, and Republican Party of India.
argument? On the basis of these survey results, which of our concepts, if any,
might be tested and refined in further research?
Before we get to the survey results, we should note that in 1971, in the elections
to the Tamil Nadu State Assembly, what we call ‘‘polity-wide’’ parties such as
Congress, or the Communist Party of India, won 55.6% of the vote. However, in
2001 ‘‘polity-wide’’ parties only won 9.1% of the total Tamil Nadu State Assembly
votes. Does this weaken our argument about ‘‘centric-regional’’ parties in state-
nation parliamentary politics? For the thirty-year decline in Tamil Nadu State
Assembly elections of polity-wide parties, see table 4.1.
Some of our survey results also demonstrate the continuing strength of the
Tamil identity. The three authors agreed that we would like to replicate for India
and Tamil Nadu the five-point scale of identity that we cited in chapter 1 for Cata-
lonia in Spain and for Flanders in Belgium. The results to this question, taken by
itself, would seem to refute some of our main arguments about the successful in-
tegration of Tamil Nadu into Indian politics. There is a much higher ‘‘only Tamil’’
response in Tamil Nadu (33.8%) than the ‘‘only Catalan’’ response in Catalonia
(11.0%) or the ‘‘only Flemish’’ response in Flanders (3.5%) (see tables 1.2 and 1.3).
136 crafting state-nations
table 4.2
Five-Point Scale of Identity in Tamil Nadu and the Rest of India, 2005
Identity Tamil Nadu Rest of India
Only regional 34 12
More regional than Indian 15 11
Equally Indian and regional 15 22
More Indian than regional 12 14
Only Indian 24 41
Source: SDSA 2005, India main dataset, weighted by state electorate, CSDS Data Unit, Delhi.
Note: ‘‘Do not know/no answer’’ responses are not included. The total number of
respondents in Tamil Nadu is 391; for the rest of India, the total number of respondents is 4,811.
Question C-16: ‘‘When we ask people as to who they are, we get different answers. Some
people say that they are only Indian, while others say they are more Indian and less regional
(e.g., Tamil). Some people say they are only Indian, while others say they are as Indian as they
are regional (e.g., Tamil). And others say they are more regional and less Indian, while others
say they are only regional (e.g., Tamil). How do you identify yourself ?’’
Moreover, when we compared the Tamil Nadu responses with those in other
Indian states, the percentage of respondents in Tamil Nadu who answered that they
were ‘‘only Indian,’’ was 17 points lower than the Indian average, and the ‘‘only
regional’’ response was 22 points higher than the Indian average (see table 4.2).
Two questions emerge from tables 4.1 and 4.2. Given the seeming collapse of
‘‘polity-wide’’ parties in the elections for the Tamil Nadu State Assembly, the
question must be asked: Are the Tamil parties no longer the ‘‘centric-regional’’
parties we showed they were in the 1971 elections? And further: if the ‘‘only Tamil’’
identity is so strong relative to other Indian states, were we wrong when we argued
that multiple but complementary state-nation-type political identities were promi-
nent in Tamil Nadu?
The first question is easier to answer. In 2004–2006, even though the polity-
wide parties had won less than 10% of the State Assembly votes in the most recent
elections, the DMK was more than ever engaged in highly rewarding ‘‘centric-
regional’’ party politics in the context of India’s parliamentary system. The DMK
and two Dravidian allies—helped by their alliance with the polity-wide Con-
gress Party—not only controlled the Tamil Nadu State Assembly but, in terms
of seats, were the third-largest party in the Congress-led ruling coalition at the
center. For this, they were rewarded with the crucial federal cabinet portfolio
of the Ministry of Finance and five of the other twenty-seven cabinet posts.51
51. See E. Sridharan, ‘‘Electoral Coalitions in 2004 General Elections: Theory and Evidence,’’
Economic and Political Weekly, vol. 39, December 18, 2004, pp. 5418–5425.
tamils in india 137
table 4.3
Levels of Trust in Central Institutions and Satisfaction with Democracy,
India and Tamil Nadu, 2005
Level of trust Tamil Nadu Rest of India
Respondents saying that they had a ‘‘great deal of trust’’
in . . .
. . . the central government 58% 30%
. . . the army 82 64
. . . the Election Commission 49 43
Respondents saying that they were ‘‘very satisfied’’ with the 35 23
way democracy works in our country
Source: SDSA 2005, India main dataset, weighted by state electorate, CSDS Data Unit, Delhi.
Note: ‘‘Do not know/no answer’’ responses are not included. Total number of respondents in
Tamil Nadu is 391; in the rest of India, 4,811. The exact N varies slightly for each question
because of some missing cases.
Questions:
C-13a: ‘‘How much trust do you have in the Central Government?’’
C-13f: ‘‘How much trust do you have in the Army?’’
C-13j: ‘‘How much trust do you have in the Election Commission?’’
C-12: ‘‘On the whole, how satisfied are you with the way democracy works in our country—
very satisfied, somewhat satisfied, somewhat dissatisfied, totally dissatisfied?’’
Indeed, a strong case can be made that for the last fifteen years, the state of Tamil
Nadu has received more influential positions at the center than any other state
in India.
The second question, about the implications of strong Tamil identity on the
state-nation project, is more complicated. It is true that a much higher percentage
of Tamils expressed an ‘‘only Tamil’’ identity than did respondents for comparable
questions in Catalonia or Flanders. But the indicator value of the identity ques-
tion is most useful to the analyst if it is part of a related bundle of questions, such as
trust in the center, or trust in a major institutions of the state such as the military,
or satisfaction with the overall quality of democracy within the polity. When we do
this, we get a more accurate indication of multiple but complementary identities.
In fact, almost twice as many Tamil respondents said they had a ‘‘great deal of trust
in the central government’’ (58%) as did the overall Indian group (30%). Also, 82%
of Tamil respondents had a ‘‘great deal’’ of trust in the Indian Army compared to
the 64% Indian average. Very importantly, 35% of Tamils stated that they were
‘‘very satisfied’’ with the way democracy works, 12 points above the Indian average.
On these results, despite their very strong ‘‘only Tamil’’ self-identification, respon-
dents in Tamil Nadu seem to confirm our hypotheses about their state-nation
integration into Indian politics (see table 4.3).
138 crafting state-nations
52. See the note to table 4.4 for how we constructed an index for positive and negative ratings.
table 4.4
Name Recognition and Evaluation of Some Major Political Figures,
Tamil Nadu and the Rest of India, 2005
Name
recognition Positive rating Negative rating
Name of leader (place of origin, social category, Tamil Rest of Tamil Rest of Tamil Rest of
and ideological orientation) Nadu India Nadu India Nadu India
Mahatma Gandhi (Hindu merchant caste from 96% 92% 79% 72% 1% 15%
Gujarat). Congress. Leader of mass mobiliza-
tion in the freedom movement. Known as ‘‘fa-
ther of the nation.’’ Founder of a new political
ideology centered around nonviolence and a
critique of modern civilization.
Source: SDSA 2005, India main dataset, weighted by state electorate, CSDS Data Unit, Delhi.
Note: ‘‘Name recognition’’ stands for percentage of respondents who answered in the affirmative to the
question about having heard the name. ‘‘Positive rating’’ stands for percentage of respondents who, from
among those who recognized the name, placed the leader from 8 to 10 on a ten-point scale. ‘‘Negative rating’’
stands for percentage of respondents who placed the recognized leader from 1 to 3 on a ten-point scale.
Question Q-59: ‘‘Now I will read out the names of Indian leaders. You tell me about these leaders one by
one, as to what was their contribution to Democracy in India. I will show you a ladder with 1 to 10 marked on
it [show card]. If 1 represents such a leader who has contributed very little to democracy and 10 represents
such a leader who has made great contribution to democracy, then where will you place [the leader]?’’
140 crafting state-nations
Throughout this book we have spoken of political identities. However, the section
we have just finished, concerning the difficulty of interpreting the political signifi-
cance of ‘‘only Tamil’’ identities, calls for explicit reflection on the uses and abuses
of ‘‘identity’’ in modern social science. Questions on identity can create a false
appearance of polarization if only two choices are given, of which the respondent
is forced to choose one. For example, a two-point-scale question, of the sort that
explicitly asks respondents to give a ‘‘primary identity’’ (when they may not in fact
want to choose between their various identities) was used in the World Values
Survey for the United States. The question itself would seem to have played an
important role in producing the following responses (see table 4.5).
We ask the reader to carry out a simple thought exercise. Assume that, instead
of the two-point scale in table 4.5, the respondents were given a five-point-scale
question of the sort we have reproduced for Spain, Belgium, and India. Is there
some chance that on a five-point scale, (only American, more American than
Latino, equally American and Latino, more Latino than American, only Latino),
the modal response among second- and third-generation Latinos would in fact be,
‘‘equally American and Latino’’?54
53. This said, we should acknowledge that on later reflection, and unfortunately too late to be
included in this round of surveys, we realized that our list had no Muslim or South Indian leaders.
In the next round of these surveys, we will certainly include a major nationalist Muslim leader such
as Maulana Azad and leaders from South India like C. Rajagopalchari, Kamaraj, and Periyar.
54. Narayani Lasala-Blanco, Ph.D. candidate at Columbia University, conducted a June 2006
survey of the two hundred most prominent Mexican-American leaders in California. From a Hunt-
ingtonian perspective, one would expect them to be the most ‘‘Mexican’’ in their self-identity.
Lasala-Blanco asked both the two-point and the five-point identity questions to the same people
during the same interview. In the two-point question, ‘‘Do you identify yourself as primarily Mexi-
can or primarily American?’’ 50% defined themselves as primarily Mexican. In contrast, when given
the choice to answer the five-point scale question, only 5% defined themselves as ‘‘Mexican only’’
tamils in india 141
table 4.5
Self-identification of Americans within a Binary Ethnic Scale
Largest historic Largest current
minority minority Majority
Self-identification (blacks) (Hispanics) (whites)
Above all, I am an American first and a 18% 22% 28%
member of some ethnic group
second.
Above all, I am a Black American. 80
Above all, I am an Hispanic American. 76
Above all, I am a White American. 65
Sources: World Values Survey: 1990–93, Ronald Inglehart et al., Inter-University Consortium for
Political and Social Research, University of Michigan, questions 208, 233. For the United States,
a cross-tabulation between questions 208 and 233 was used.
Note: Question 208 is the following: ‘‘Which of the following best describes you?’’ The
columns do not add up to 100 because of ‘‘do not know’’ and some isolated ‘‘other’’ answers.
But our excursus needs to go even further. The data we have produced about
#5, ‘‘only Tamil’’ responses, for Tamil Nadu, raise another set of concerns of the
sort that need to be more systematically incorporated into studies of identity—
especially of identities with some presumed political implications. Is it correct to
assume that respondents who express very strong ethnic, regional, or national
identity, such as ‘‘only Tamil,’’ would have a lower degree of trust in the central
government, or a lower degree satisfaction ‘‘with the way democracy works’’ than
the polity-wide average? We designed our battery of questions to shed light on
such questions. In the case of Tamil Nadu, our results distinctly call into doubt the
index value of responses to identity questions, taken in isolation. In fact, those
respondents who self-identified as ‘‘only Tamil’’ were more than twice as likely
(62.6%) as the all-India average (29.5%) to express a ‘‘great deal of trust in the
Central Government.’’ The ‘‘only Tamil’’ respondents were also above the all-
India average in saying that they were ‘‘very satisfied with the way democracy
works in our country.’’ For the complete results and breakdown, which show that
no matter which of the five self-identities they chose in Tamil Nadu, all five of
them are at or substantially above the all-India average, see table 4.6.
and only 15% identified themselves as ‘‘more Mexican than American.’’ Thus, 95% of them included
American as part of their identity. Indeed, the modal self-identification was ‘‘equally Mexican and
American.’’ See Lasala-Blanco, ‘‘Who Are ‘They’? The Real Challenges of Mexican Immigration,’’
paper presented at the 2006 Midwest Association of Public Opinion Research Conference, Novem-
ber 17–18, 2006, Chicago, Illinois.
142 crafting state-nations
table 4.6
Identity by Level of Trust in Central Government and Satisfaction
with Democracy in India and Tamil Nadu, 2005
Those in Tamil Nadu who identify themselves as:
All-India Only More Tamil Equally Indian More Indian Only
Those who say they . . . average Tamil than Indian & Tamil than Tamil Indian
table 4.7
National and Regional Pride in India and Tamil Nadu, 2005
Pride Tamil Nadu Rest of India
Very proud of state or regional identity 67% 49%
Very proud of being Indian 73 65
Source: SDSA 2005, India main dataset, weighted by state electorate, CSDS Data Unit, Delhi.
Note: Total number of respondents in Tamil Nadu is 391; the rest of India, 4,811. The exact N
varies slightly for each question due to some missing cases.
Questions:
48: ‘‘How proud are you to be a [regional identity name, e.g. Tamil]—very proud, proud, not
proud, or not at all proud?’’
49: ‘‘How proud are you to be an Indian—very proud, proud, not proud, or not at all proud?’’
‘‘very proud of being Indian.’’ Indeed, they were somewhat more proud of being
Indian than the all-India average; 73% versus 65%. Pride, like trust, can be a
multiple-sum rather than a zero-sum relationship (see table 4.7).55
The whole ideology of self-determination and nationalism that undergirds
nation-state theory and the democratic conception of ‘‘let the people decide’’
means that solving the nationality conflict normally entails the holding of a
plebiscite wherein people decide between two mutually exclusive alternatives.
Such formulations ignore the fact that many people might have dual identities
and may not want to make what seems to them a dichotomous decision that
precludes living in a political community we would call a state-nation.
55. For the overlap, or non-overlap, between identity and attitude toward independence in the
Basque Country in Spain, see the analysis and the table in Juan J. Linz, Conflicto en Euskadi
(Madrid: Espasa Calpe, 1986), pp. 136–148.
chapter five
The literature on nationalism is filled with warnings about the ‘‘slippery slope of
ethno-federalism,’’ which in the case of Yugoslavia (but not in India, Belgium, and
Spain) we are in substantial agreement.1 However, much less studied and equally
prevalent is the ‘‘slippery slope toward violence and secession’’ in countries that
have employed aggressive nation-state building policies in societies that are multi-
national. We hope that this chapter is a contribution to that undertheorized issue.
There are three issues concerning Sri Lanka about which there is a domestic and
international consensus.2 First, even though about 70% of the population was
Sinhalese (92% of whom were Buddhists and almost all of whom spoke Sinhalese)
and 23% were Tamil (about 86% of whom were Hindus and most of whom spoke
Tamil) at Independence in 1948, there had been no riots or any other form of
collective violence between Sinhalese and Tamils for hundreds of years before
Independence.3
1. The most influential book cited to advance this argument is Bunce, Subversive Institutions:
The Design and the Destruction of Socialism and the State (New York: Cambridge University Press,
1999). The ‘‘subversive institutions’’ of the title are those institutions with some ethno-federal dimen-
sions that deviate from the classic nation-state model.
2. Until 1972 Sri Lanka was called Ceylon. In this chapter we will use the names interchange-
ably, with the aim of achieving historical and contextual appropriateness.
3. The only major ethnic riot had been between Sinhalese and Muslims in 1915. Muslims, who
in 1946 constituted 6% of the population, were predominately Tamil-speakers and thus on many, not
tamils in sri lanka 145
all, policy issues of language often allied with the Hindu Tamil-speakers. The population estimates
are from the 1946 Census as reported, with more data on territorial concentrations of different
categories of Sinhalese, Tamils, and Muslims in Chandra Richard de Silva, ‘‘Sinhala-Tamil Rela-
tions and Education in Sri Lanka: The University Admissions Issue—The First Phase, 1971–77,’’ in
From Independence to Statehood: Managing Ethnic Conflict in Five African and Asian States, ed.
Robert B. Goldman and A. Jeyaratnam Wilson (London: Francis Pinter, 1984), table 9.1, p. 136. For a
number of tables on Ceylon’s religious, linguistic, and ethnic composition at independence, see
Robert N. Kearney, Communalism and Language in the Politics of Ceylon (Durham: Duke Univer-
sity Press, 1967), pp. 1–18.
4. See Obeyeskere, ‘‘Origins and Institutionalization of Political Violence,’’ in Sri Lanka in
Change and Crisis, ed. James Manor (New York: St. Martin’s Press, 1984), pp. 153–174, quotation
from p. 153, emphasis added.
5. See Tambiah, Sri Lanka: Ethnic Fratricide and the Dismantling of Democracy (Chicago:
University of Chicago Press, 1986), esp. pp. 13–64, quotation from p. 7.
6. Donald L. Horowitz, ‘‘Incentives and Behavior in the Ethnic Politics of Sri Lanka and
Malaysia,’’ Third World Quarterly 11 (October 1989), p. 18. His monumental book on about 150
ethnic conflicts is Ethnic Groups in Conflict (Berkeley: University of California Press, 1985).
7. See Howard Wriggins, Ceylon: Dilemmas of a New Nation (Princeton: Princeton University
Press, 1960), quotations from pp. 328 and 282, respectively.
146 crafting state-nations
East Asia, Ceylon has more of the attributes of a modernized social and political
system than any other.’’ These attributes included 60% literacy at the time of
Independence—higher than India in 2000. Ceylon also had ‘‘the highest per
capita income of any country in Asia except for Japan.’’ Using contemporary
vocabulary, we would say that Ceylon at Independence had a much more ‘‘use-
able state’’ than most developing countries. For, as Wriggins argues, the civil
service by 1949 was ‘‘almost exclusively Ceylonese. The public service has always
had a high reputation for integrity and the impersonal application of the law.’’
Ceylonese society was also admirably progressive: ‘‘Important social welfare pro-
grams accounted on the average for over 35 per cent of the government’s expendi-
ture for the first ten years of independence,’’ which was almost twice the amount
of the budget India spent on social welfare. Indeed, Ceylon was considered such
an outstanding social model of development that it was ‘‘used as a training ground
for South Asian administrators to study public health administration . . . low cost
housing and the development of cooperatives.’’8
The third consensus is that there were major Sinhalese/Tamil riots in 1958,
1977, 1981, and 1983, and that since 1983 Sri Lanka was in the grip of a brutal
internal war of secession over the independence of the Tamils. Meanwhile, the
quality of democracy, once seen as one of the best in Asia, deteriorated—indeed,
at times, broke down.9 The magnitude of the breakdown may surprise some:
according to one of the standard compilations of deaths by political violence, it
appears that for the 1946–2007 period, the Sinhalese/Tamil civil war in Sri Lanka
contributed to the highest per capita death rate among democracies or near-
democracies in the world.10 The death rate intensified in the (Sinhalese-led) army
assault of 2008–2009 against the (Tamil-led) LTTE insurgents, resulting in an
estimated twenty thousand deaths.11
Given these three patterns, it is clear that we cannot explain Sri Lanka’s recent
ethnic conflicts by recourse to arguments about ‘‘historic hatreds,’’ ‘‘statelessness,’’
or ‘‘absolute poverty.’’ This, then, presents a puzzle: Why did the non-issue of
Tamil separatism become the issue? What explains the outbreak of unprece-
dented violence in Sri Lanka a decade after Independence?
In our chapter on Tamil Nadu, we analyzed how an issue became a non-issue.
In this chapter, we will attempt to analyze and explain how a potential issue
became a civil war. In terms of the overall framework we developed in the first
chapter, Ceylon at Independence was a society with some multinational dimen-
sions. But it was not, as India was at Independence and Spain and Belgium were
in the 1970s, a ‘‘robustly multinational’’ polity. We say this because there were no
significant political leaders in the territory of the new state who were seeking
independence from Ceylon.
One of the many reasons that no important Tamil leaders argued for indepen-
dence in the 1940s was that the Tamil population was physically divided into three
quite separate territories. The populations in each of these three territories had sig-
nificantly different cultures, language traditions, employment patterns, and, from
the orientation of rational choice theory, preferences, concerning their needs as
‘‘Tamils’’ in Ceylon. One group of Tamils have been concentrated in the northern
peninsula of Ceylon, where they established the sixteenth-century Tamil Hindu
kingdom of Jaffna in which they were a linguistic, religious, and ethnic majority.
Many Tamils from this area eventually settled in Colombo, the capital of Ceylon,
located in the country’s southwest. This group included many Tamil civil servants,
lawyers, merchants, and workers. In the historical literature about Ceylon, these
two groups of Tamils are referred to as ‘‘Ceylon Tamils.’’ The surveys we utilize
later in this chapter refer to these two groups as ‘‘Tamils.’’ Therefore, for simplic-
ity’s sake, we will also refer to these two groups as ‘‘Tamils.’’ A third territorially
concentrated group of Tamils exists in Ceylon/Sri Lanka. This group began com-
ing from India as laborers in the mid- to late nineteenth century to work on the new
tea plantations in the central highlands of Ceylon. In the historical literature,
these later-arriving, plantation-based Tamils are called ‘‘Indian Tamils.’’ In this
book, and in the surveys we use, they are referred to as ‘‘Up-Country Tamils.’’ At
Independence, the Up-Country Tamils, and the Tamils around Colombo—who
constituted approximately half of all the Tamils in Ceylon and lived in Sinhalese-
speaking majority districts—were unenthusiastic about purely territorial responses
to Tamil needs such as federalism, much less independence.
Just as important, as we shall soon document, Tamils at Independence, be-
cause they had a tradition of strong university education in English, were well
148 crafting state-nations
12. For a review of non-territorial proposals to protect their interests advanced by the Tamils
before Independence, see Wilson, Break-Up of Sri Lanka, pp. 1–24.
13. See Wilson, Break-Up of Sri Lanka, p. 86, emphasis in original.
14. See figure 1.1. Our position is that after 1983, Sri Lanka’s polity increasing was in the upper
right hand circle, the position we argued was most difficult and unlikely for peace and unity in one
democratic state.
tamils in sri lanka 149
building policies in Sri Lanka put the multinational society on the ‘‘slippery slope’’
of state erosion and democratic decay. Let us examine five major steps down this
slippery slope.
Let us analyze five different but compounding political choices that helped de-
construct Ceylon’s multicultural and peaceful polity and construct two warring
nations in one territory.
1. The disenfranchisement of the ‘‘Up-Country Tamils’’ in 1947–48. The Up-
Country Tamils had voted in all of the general elections in Ceylon since the
securing of universal suffrage in 1931. The Up-Country Tamils constituted just
over half of the entire Tamil population in Ceylon.15 By 1948, however, they had
been disenfranchised by the Sinhalese-dominated government. The justification
for this disenfranchisement advanced by the government in the parliamentary
debates was that many of the Up-Country Tamils had come from South India to
work in the tea plantations in the late nineteenth century and were not really part
of the Ceylonese nation. Thus, they should not be voters in Ceylon. In fact, many
of the Up-Country Tamils were second- and third-generation residents in Ceylon
and had severed most of their ties with India.
The Up-Country Tamils were situated in central Ceylon in the hill country
and were physically separated from the more educated Tamils in the north around
Jaffna, the Jaffna Zone Tamils. For this reason and others, Up-Country Tamils
developed their own political associations. Their party, the Ceylon Indian Con-
gress, won seven seats in the general election of 1947.16 More important, they
were a substantial minority in many other constituencies, forming coalitions with
polity-wide non-Tamil parties, especially the Marxists. Sir Ivor Jennings, the emi-
nent British constitutionalist and drafter of the Ceylonese constitution, estimates
that the Up-Country Tamils were ‘‘decisive’’ in the election of twelve to fourteen
Marxist candidates.17 Since the All Ceylon Tamil Party won seven seats in the
north, Tamils thus had a significant impact on about twenty of the ninety-one
15. See Amita Shastri, ‘‘The Tamil Citizenship Act of 1948 and Sri Lanka Partition,’’ Contempo-
rary South Asia 8 (March 1999), pp. 25–86.
16. See Kearney, Communalism and Language, p. 104.
17. Sir Ivor Jennings, The Constitution of Ceylon, 3rd ed. (London: Oxford University Press,
1953), p. 29.
150 crafting state-nations
constituency races for the first parliament. In this context, the leader of the All
Ceylon Tamil Party, G. G. Ponnambalam, argued for ‘‘constructive collaboration’’
and allied his party with the governing United National Party (UNP). For this
support, Ponnambalam was made minister for industries.
We should note that the disenfranchisement had three consequences. First, by
disenfranchising the Up-Country Tamils in the name of nation-state ethnic au-
thenticity, the ruling centrist party of course conveniently eliminated the electoral
ally that had been ‘‘decisive’’ for the victory of at least twelve ethnically ‘‘pure’’
Sinhala, but politically ‘‘impure’’ Marxist, party seats. Second, after the disen-
franchisement, the Tamil parties never again had such a high percentage of seats
in parliament and never again formed so many electoral alliances with winning
(Marxist or non-Marxist) polity-wide non-Tamil parties. Third, and most impor-
tant, their disenfranchisement contributed to the number two leader of the All
Ceylon Tamil Party, S. J. V. Chelvanayakam, leaving this polity-wide party on the
argument that the Tamils needed a territorial electoral base in the north to protect
Tamil interests. He founded the Federal Party, which, after more rebuffs of Tamil
‘‘recognition’’ that we shall analyze later, increasingly became the political base
for Tamil demands for some form of territorial autonomy.
The disenfranchisement also had an effect on competitive politics within the
Sinhalese/Buddhist majority. With the reduced coalitional capacity of the Tamils,
the temptation increased for winner-take-all outbidding among Sinhalese parties
for the votes of Sinhalese for the control of their nation-state. It was not inevitable,
however, that such outbidding would occur, or, if it did occur, that it would be
successful.
2. The 1956 ‘‘Sinhalese Only’’ electoral campaign and the marginalization of
polity-wide careers for Tamil- and English-speakers. The political leader who initi-
ated the appeals to aggressive nation-state policies was S. W. R. D. Bandaranaike.
Born into an aristocratic family, Bandaranaike was raised as an English-speaking
Christian. A brilliant orator, he first won recognition in Ceylon by being elected
secretary of the Oxford Union in 1923. He converted to Buddhism upon his return
to Ceylon and, by the mid-1930s, combined the portfolios of the ministries of local
government and health and was later a minister in the first post-Independence
cabinet. He used this power base to increase his personal following by appealing
to problems involving the Sinhalese language and of Buddhists. Language and
religion are of course highly interconnected, but let us first begin with an analysis
of the ethnic mobilization of language and its divisive effects.
It is crucially important to stress that Bandaranaike initially found scant sup-
tamils in sri lanka 151
18. For his biography, see James Manor, The Expedient Utopian: Bandaranaike and Ceylon
(Cambridge: Cambridge University Press, 1989), quotation from p. 201. The government wanted to
introduce more Sinhalese and Tamil gradually, but Bandaranaike wanted them adopted as the
‘‘official languages immediately.’’ Wriggins, Ceylon, p. 121.
19. On the ‘‘no compete’’ agreement and the language issue in the 1956 campaign, see Kearney,
Communalism and Language, pp. 68–89.
152 crafting state-nations
the 1956 campaign revolved around the extreme nationalist competition of the
two largest parties for, in effect, nationalist Sinhalese votes. Bandaranaike won
with a strong parliamentary majority.
The bill to make Sinhalese the ‘‘one official language’’ of Ceylon was intro-
duced on June 14, 1956, and was passed nine days later. Four sentences in the fine
book on Ceylon by Wriggins capture the tone of the aggressive nation-state policy
making process that followed.
The polarization in the campaign was such that neither the former ruling
party, UNP (which had won 40% of the seats in the northeast in 1952), nor the new
ruling party, SLFP, contested a single constituency in this Tamil-speaking zone in
1956. For their part, the Tamil parties, especially the Federal Party, limited their
campaigning almost entirely to Tamil-speaking areas.21 In our theoretical frame-
work, this means that Ceylon ceased to have any major polity-wide parties after
the aggressive nation-state-building policies were initiated in 1956.
In our conceptual framework, we also argued that given the inherent tensions
in a multinational society, polity-wide jobs in the state apparatus for all citizens are
normally useful for polity-wide integration. The ‘‘Sinhalese Only’’ nation-state
policies eroded this source of Tamil integration and identification with the Cey-
lonese political community.
After the laws on language of 1956, Tamil-speakers could still take the higher
civil service entrance exam in English but, after admission, further tests for pro-
motion would be conducted in Sinhalese only. Existing Tamil-speakers would
have to show proficiency in Sinhalese by 1960. English was abolished for the
liberal arts at the university level. Tamils had been overrepresented in public
employment, but the rapidity in the drop of their numbers in the state sector
caused growing alienation in this politically critical group.
In 1955, 26% of the members of the elite and politically powerful Sri Lankan
administrative service (SAS) were Tamils. By 1979 this figure had been cut in
half.22 But for younger Tamil aspirant professionals, recruitment trends into the
SAS were even more ominous. From 1970–77, of the 467 admissions via the Open
Competitive Examination route, 34 (only 7.2%) were Tamils. In the 1977–81
period, of the 159 new admissions via this route, none were Tamils.
A new set of admissions criteria to universities also reduced Tamil chances of
getting professional qualifications for private sector jobs. This was particularly so
in engineering and medicine, where Tamils had been traditionally strong. In
1969, 48% of the new admissions in engineering, and 49% in medicine, were
Tamils. In 1974, these figures had dropped to 16% and 26%, respectively.23
Concerning primary and secondary education, the principal goals of nation-
state policies were to eliminate English and to make it compulsory for all children
of Sinhalese parents to learn Sinhalese. Children of Tamil parents were taught
Tamil but were often not provided with teachers to learn Sinhalese. Thus, iron-
ically, this nation-state policy was for ‘‘Sinhalese Only.’’
3. The mobilization of religious differences. A potential problem of harmony in
Ceylon’s multiethnic but not yet ‘‘robustly multinational’’ polity that needed ad-
dressing at Independence was the status of the Buddhist community. Smith’s
standard account of religion and politics in South Asia in the postcolonial era
opens its discussion of Buddhism with the following assessment of the colonial
legacy: ‘‘The Sinhalese Buddhist majority was in a markedly inferior position vis-
à-vis other communities, and its language, religion and culture had been rele-
gated to a secondary place.’’24 This is true. At Independence, Buddhist monu-
ments, which had received very little support during the period of colonial rule
from the Catholic power Portugal, the Dutch Protestants, or from the successor
Anglican/secular power Britain, were in a sad state of disrepair. Also, during the
British period, much of the public expenditures for schooling went to ‘‘state-
assisted’’ English-language privately run schools. These schools, the best in Cey-
lon, often had British and American missionary origins. They were attended
22. These data are from the official publication, Ceylon Civil List, and were reported in S. W. R.
de A. Samarasinghe, ‘‘Ethnic Representation in Central Government Employment and Sinhala-
Tamil Relations in Sri Lanka: 1948–81,’’ table 11.3, p. 177.
23. See C. R. de Silva, ‘‘The Impact of Nationalism on Education: The Schools Takeover (1961)
and the University Admissions Crisis, 1970–1975,’’ in Collective Identities, Nationalisms, and Protest
in Modern Sri Lanka, ed. Michael Roberts (Colombo, Sri Lanka: Marga Institute, 1979), pp. 474–
499, data from table 8.
24. Donald E. Smith, ‘‘The Sinhalese Buddhist Revolution,’’ in South Asian Politics and Reli-
gion, ed. Smith (Princeton: Princeton University Press, 1966), p. 454.
154 crafting state-nations
25. Among the most important works are Richard Gombrich and Gananath Obeyesekere,
Buddhism Transformed: Religious Change in Sri Lanka (Princeton: Princeton University Press,
1988); Gananath Obeyesekere, ‘‘Buddhism, Nationhood, and Cultural Identity: A Question of
Fundamentals,’’ in Fundamentalisms Comprehended, ed. Martin E. Marty and R. Scott Appleby
(Chicago: University of Chicago Press, 1995), pp. 231–258; Stanley J. Tambiah, ‘‘Buddhism, Politics,
and Violence in Sri Lanka,’’ in Fundamentalisms and the State: Remaking Politics, Economies, and
Militance, ed. Martin E. Marty and R. Scott Appleby (Chicago: University of Chicago Press, 1993),
pp. 589–619; and Heinz Bechert, ‘‘S.W.R.D. Bandaranaike and the Legitimation of Power through
Buddhist Ideals,’’ in Religion and Legitimation of Power in Sri Lanka, ed. Bardwell L. Smith
(Chambersburg, PA: Anima Books, 1978), pp. 199–211. Also useful, especially because it implicitly
introduces an Indian perspective, is Urmila Phadnis, Religion and Politics in Sri Lanka (New Delhi:
Manohar, 1976).
26. Obeyesekere, ‘‘Buddhism, Nationhood, and Cultural Identity,’’ p. 233.
27. Ibid., p. 236.
28. Ibid., p. 254.
29. Ibid., p. 247.
tamils in sri lanka 155
Buddhist revivalism stimulated and sponsored by Colonel Olcott and the Bud-
dhist Theosophical Society [he] founded in 1880 was the establishment of Bud-
dhist schools to counter the near monopoly that the Protestant missions (and to a
lesser extent the Catholic Church) had over the educational system.’’30 Olcott also
contributed to a tradition of charismatic younger monks such as Dharmapala
taking up Buddhist mobilization directly.
But there is no direct line between this period and the 1956 events we have
described. In point of fact, this brand of Buddhist nationalism went into abeyance.
As Tambiah writes, ‘‘A remarkable feature of the Buddhist fundamentalist and
Sinhala movement is [that in 1915 it] seemed to lose prominence and surrendered
the lime light to a different cast of Sinhalese and Tamil politicians, who were to
initiate a phase of collaboration rather than confrontation.’’31
The 1947 election witnessed another period of Buddhist involvement in nation-
alist politics in which monks (bhikkus), particularly younger Marxist-influenced
monks, created a party, the Ceylon Union of Bhikkus (LEBM), that played a major
mobilizing role in the election. However, this movement also rapidly lost momen-
tum partly because the young bhikkus were too leftist for the mainstream parties
(particularly the ruling, center-right UNP party) and too radical, too young, and
too autonomous for the senior monks traditionally in charge of the Sangha. As
Tambiah writes, ‘‘Once the election of 1947 was over and the UNP was elected,
[the LEBM] soon became defunct.’’32 The Indian scholar Urmila Phadnis concurs
that ‘‘the LEBM virtually became defunct once the 1947 election was over. Its
leadership lost its initiative, and the group broke up.’’ She goes further, establish-
ing that religious issues themselves lost salience in Ceylonese public debates. In
fact, she makes the surprising assertion that ‘‘in such a setting, religious issues were
relegated to a peripheral position, if at all in the 1952 election. In fact, several
parties ignored the subject altogether in their manifestos.’’33
The above discussion underscores our central point, that religion only emerged
as the critical dividing point in the Ceylonese polity once a major party leader
found a way to mobilize it in order to ride into state power by polarizing the
majority Sinhala/Buddhists against the Tamil/Hindus. S. W. R. D. Bandaranaike,
as we have already shown, did precisely this to win the 1956 election. As Tambiah
writes, the leadership was not from within the Sangha; political bhikkus had
to wait to ‘‘join a more congenial political coalition under the leadership of
such as information technology. But when Ceylon began addressing the problems
of the Sinhalese-Buddhist majority, political leaders went from relative neglect to
aggressive nation-state policies, which contributed to increasingly polarized polit-
ical identities and eventually to war between the two nations.
Let us look specifically at the role of religion in the years leading up to the 1956
election. Bandaranaike knew he had a powerful conjunctural factor in his favor.
The election year coincided with the great celebrations planned to commemorate
the 2500th anniversary of the death of Buddha and the first landing of his followers
in Ceylon. Bandaranaike initially received quite lukewarm support from impor-
tant parts of his Buddhist electoral audience. Wriggins writes that Bandaranaike
in 1951 ‘‘courted Buddhist opinion by protesting against the government’s neglect
of the Buddhist religion and culture. At the time he publicly proposed making
Buddhism a state religion, but both Buddhist priests and laymen objected.’’37
Indeed, as late as January 1956, the year of the fatal mobilization, the president of
the All Ceylon Buddhist Congress, Dr. G. P. Malalasekera, in a major speech
about the forthcoming report on the status of Buddhism, explicitly argued that
Buddhists ‘‘have no desire to make Buddhism the State religion—in spite of the
cry raised by self-seeking politicians—but they do want the State to help them
rehabilitate themselves and undo some, at least, of the injustices perpetrated
against them during the days of their [colonial] subjection.’’38
After his 1956 electoral victory, Bandaranaike initially concentrated on passing
the polity-fragmenting majoritarian language policies we have discussed. How-
ever, the political promotion of identification of the state with only one, of what
were then two, self-conscious political communities in Sri Lanka (and four signifi-
cant religions) was deepened in 1978 when the new constitution declared, ‘‘The
Republic of Sri Lanka shall give Buddhism the foremost place and accordingly it
shall be the duty of the State to protect and foster the Buddha Sasana’’ (Article 9).
As the comparativist Donald Horowitz put it, ‘‘The Sri Lankan language and state
religion provisions symbolically wrote the Sri Lankan Tamils out of the polity.’’39
4. The privileging of nation-state over state-nation values. Once these nationalist
forces were ushered into existence by political leaders in the 1956 campaign and
Sinhalese/Buddhist values and language were accorded primacy, even a major
pact by Prime Minister Bandaranaike himself and the Tamil president of the
37. Wriggins, Ceylon: Dilemmas of a New Nation (Princeton: Princeton University Press, 1960),
p. 121. This particular speech was given at the Young Men’s Buddhist Association.
38. Text of speech reproduced in The Times of Ceylon, January 15, 1956, cited in Wriggins,
Ceylon, p. 196.
39. Horowitz, ‘‘Incentives and Behavior in the Ethnic Politics of Sri Lanka and Malaysia,’’ p. 28.
158 crafting state-nations
40. We are referring to the Bandaranaike-Chelvanayagam Pact of July 1957 and the Dudley
Sennayake–Chelvanayagam Pact of March 1965. Both pacts are reproduced in Ambalavanar Sivara-
jah, Politics of Tamil Nationalism in Sri Lanka (New Delhi: South Asian Publishers, 1996), pp. 202–
205. Many analysts and policy makers, Sinhalese and Tamil alike, now argue that if either pact had
been implemented, the civil war would not have happened.
41. Wriggins, Ceylon, p. 270.
42. Tambiah, Sri Lanka, p. 26.
43. See Newton Gunasinnghe, ‘‘Community Identity and Militarization in Sri Lanka: Sri Lan-
kan Armed Forces,’’ in The Challenge in South Asia: Democracy, Development, and Regional
Cooperation, ed. Ponna Wijnaraja and Akmal Hussain (New Delhi: Sage Publications, 1989); and
Bose, States, Nations, and Sovereignty, p. 75. See also Darina Rajasingham-Senanayake, ‘‘Sri Lanka:
Transformation of Legitimate Violence and Civil-Military Relations,’’ in Coercion and Governance:
The Declining Political Role of the Military in Asia, ed. Muthiah Alagappa (Stanford: Stanford
University Press, 2001).
tamils in sri lanka 159
Supreme Court and Court of Appeals’’ (article 151). On top of all this, the presi-
dent may, with some minor limitations, ‘‘from time to time, by Proclamation
summon, prorogue, and dissolve Parliament’’ (article 70).
Over the next decade, the president called a referendum under dubious cir-
cumstances, extended the life of the parliament (where he had a massive major-
ity) to six years longer than the prescribed term of office, passed some draconian
security laws, and replaced most judges.44 Such a powerful presidential office
makes it virtually impossible for the top political office ever to be a serious, chief-
executive-binding coalition. It also makes it extremely difficult for that office to be
a ‘‘divisible good,’’ with many of the groups, or parties, in a multinational society
sharing power, as is possible in a multinational society with a parliamentary
system. As we noted in chapter 1, all the democracies in the world that are close to
the state-nation ideal type are essentially parliamentary and thus have power-
sharing potential.
5. The elimination of the major Tamil party from parliament and the shift of
Tamil power to secessionist guerrillas. Since 1953, in sharp contrast to the pattern
we documented for Tamil Nadu, no politician ever elected to parliament from the
Tamil base in the north was ever made a cabinet minister. However, they still
contested parliamentary seats and participated in parliament. Even this ended in
1983, mostly due to nation-state policy choices.
In May 1976, twenty years after the crises had begun, most of the Tamil parties
united to create the Tamil United Liberation Front. At their founding meeting
they resolved: ‘‘The restoration and reconstitution of the Free, Sovereign, Secular,
Socialist State of Tamil Eelam based on the right of self determination inherent in
every nation has become inevitable in order to safeguard the very existence of the
Tamil nation in this country.’’45
Many Sinhala observers saw this as a Declaration of Independence. However,
despite the provocative language, one of the chief Tamil negotiators, A. Jaya-
ratham Wilson, writing in 1978, said that the resolution ‘‘still left room for ma-
noeuvre on the question of autonomy or sovereign statehood.’’46 Indeed, in a later
44. For a discussion of the president’s extensive use and abuse of these powers, see C. R. de Silva,
‘‘Plebiscitary Democracy or Creeping Authoritarianism? The Presidential Election and Referen-
dum of 1982,’’ M. P. Moore, ‘‘The 1982 Elections and the New Gaullist-Bonapartist State in Sri
Lanka,’’ and Priya Samarakone, ‘‘The Conduct of the Referendum,’’ all in Sri Lanka in Change
and Crisis, ed. Manor, pp. 35–50, 51–75, and 76–83, respectively. For a critical analysis of the
1978 constitution by a leading Tamil scholar who frequently represented the Tamil leader, Chel-
vanaykam, in major negotiations, see A. Jeyaratnam Wilson, The Gaullist System in Asia: The
Constitution of Sri Lanka (1978) (London: Macmillan, 1980).
45. Wilson, Break-Up of Sri Lanka, p. 89.
46. A. Jeyaratnam Wilson, Politics in Sri Lanka, 1947–79 (London: Palgrave/Macmillan, 1974),
160 crafting state-nations
book, Wilson devotes most of his discussion of the 1976–83 period to negotiations
between the government and the Tamils about proposals for devolution, in which
the Tamil political leaders were apparently still quite interested.47
In 1983, the worst riots against the Tamils occurred. Shortly afterward, the
president pushed through the Sixth Amendment to the constitution, which re-
quired all members of parliament to swear a public oath against secession—and,
implicitly, for the continued unitary status of Sri Lanka—or to lose their seat in
parliament. From a comparative perspective, it is important to stress that parties
that advocate independence, such as the Scottish Nationalists in the United
Kingdom, the Parti Québecois in Canada, and the Vlaams Belang in Belgium are
able to remain in their respective legislatures. As it was, not a single Tamil mem-
ber of parliament from the north took the oath in Sri Lanka in 1983; they all
walked out of parliament, and they never returned during the civil war.48 Nu-
merous observers considered this the end of any possibility of peace and democ-
racy between Sinhalese and Tamils in Sri Lanka.
One of the Tamil members of parliament who refused to take the oath insists
that, until the riots and the oath in 1983, ‘‘the vast majority of the Tamil parliamen-
tary group would have accepted a serious federal alternative. Even outside the
Parliament, we were still the most powerful group. But, by 1985–86, the LTTE
[the separatist guerrillas] were the most powerful group.’’49 Some of the former top
guerrilla leaders agree with this assessment. In the early 1980s there were five
small guerrilla groups in the jungle: their acronyms were LTTE, PLOTE, EPRL,
EROS, and TELOS. In an interview, a former leader of EPRL estimated that
before the 1983 pogrom and oath, they ‘‘were recruiting but had no fighters—but
by 1985, they had a cadre of seven thousand.’’ In his estimation, before 1983 ‘‘most
Tamils still probably had more faith in their elected officials in Parliament than in
p. 154. A participant at the meeting, Siva Sitaparam, told Stepan in an April 1998 interview in
Colombo that a lot of politicians at the meeting had reservations about the resolution, but he
had become convinced, after a conversation with the key Tamil leader at the meeting, C. V.
Chelvanayakam, that ‘‘C. V. was of the view that if a good alternative inside Sri Lanka emerged he
would consider it.’’ Sitapuram at the time of the interview was the secretary general of what
remained of TULF.
47. For his detailed discussions of various devolutionary schemes (which we consider quite
modest by the standards of Canadian, Spanish, Belgium, or Indian devolution), see Wilson, Break-
Up of Sri Lanka, pp. 141–174.
48. The oath, in its entirety, was: ‘‘I will uphold and defend the Constitution of the Democratic
Socialist Republic of Sri Lanka and that I will not, directly or indirectly, in or outside Sri Lanka,
support, espouse, promote, finance, encourage or advocate the establishment of a separate state
within the territory of Sri Lanka.’’ See the Seventh Schedule, Article 157, of the 1978 Sri Lankan
constitution.
49. Previously cited interview by Alfred Stepan with Siva Sitaparam, Colombo, April 1998.
tamils in sri lanka 161
the boys in the jungle.’’ A former top leader of PLOT estimates that before 1983,
none of the five guerrilla groups had more than thirty ‘‘militants’’ but that by 1985,
they had four thousand under arms.50
Traumatic events can constitute a strong feeling of ‘‘we-feeling,’’ or even of a
hostile ‘‘nation,’’ among a previously loyal and quite divided minority community.
Multiple but complementary identities can be shattered and then replaced by
conflictual polar identities. In the judgment of Tambiah, policies ‘‘implemented
in the name of nation building [Sinhalese] and equitable democratization of
society served to deepen the Sinhalese-Tamil rift and increasingly to politicize
and make collective adversaries out of the Sinhalese and the Tamils.’’51 Bose
strikes a similar note: ‘‘Nation building, as attempted by those at the helm of the
postcolonial Sri Lankan state, actually facilitated the emergence and develop-
ment of a Tamil national consciousness and eventually precipitated a crisis of the
state.’’52 Let us now examine the state of opinion in Sri Lanka in 2005.
In 2005, the State of Democracy in South Asia (SDSA) project of the CSDS
carried out a survey with 4,613 respondents in Sri Lanka. In addition to the set of
SDSA questions common to all of the five South Asian countries (India, Pakistan,
Bangladesh, Sri Lanka, and Nepal), the survey included a special battery of
questions designed for Sri Lanka. In the next few pages, ‘‘Tamils’’ refers only to
‘‘Jaffna-area Tamils,’’ not to ‘‘Up-Country Tamils.’’ The responses to these ques-
tions shed great light on Sinhala/Tamil relations after fifty years of aggressive
nation-state building policies in Sri Lanka. The results dramatically reveal the
majoritarianism among the Sinhala/Buddhist population and the sense of insecu-
rity and low trust in key state institutions among the Tamils of Sri Lanka.53
For example, only 9% of Sinhala respondents ‘‘strongly disagreed’’ with the
50. Both the interviews with former guerrilla leaders were by Stepan in Sri Lanka, February
2003. A former AFP correspondent who did over one hundred interviews with guerrillas from 1987
to 1989 estimates that none of the five major guerrilla groups had more than 50 active members in
1981 but that ‘‘the July 1983 riots opened a floodgate of young Tamils to various Tamil militant
groups. It was a remarkable change in the Tamil heartland. Until then not many had vowed to
plunge into militancy.’’ See M. R. Narayan Swamy, Tigers of Lanka: From Boys to Guerrillas (Delhi:
Konark Publishers, 1994), p. 96.
51. Tambiah, Sri Lanka, p. 75.
52. Somantra Bose, ‘‘State Crises and Nationalities Conflict in Sri Lanka and Yugoslavia,’’
Comparative Political Studies 28 (April 1995), p. 100.
53. Parts of northern and eastern provinces where civil war prevented a survey were excluded
from the sampling frame, which did not cause undersampling of Tamils (11% of our sample) and Up-
Country Tamils (4%).
162 crafting state-nations
table 5.1
Perceptions of Discrimination by Community in Sri Lanka, 2005
Up-country
Respondents who: Sinhalas Tamils Tamils Muslims
Strongly disagree that ‘‘everyone enjoys 9% 32% 18% 29%
equal rights.’’
Strongly disagree that ‘‘people are free to 6 25 19 15
speak their mind without fear.’’
Say they feel ‘‘less safe’’ now ‘‘compared 20 41 21 34
to the situation in this city/town/
village a few years ago.’’
Do not feel or are unsure that, ‘‘if you 51 73 71 45
went to the police, you would be
treated in the same way as everyone
else.’’
Source: SDSA 2005, CSDS, Delhi, Sri Lanka dataset; questions A-10, C-20b.
Note: The overall N for SDSA in Sri Lanka is 4,613. In tables 5.2 to 5.5, the N varies for each
question because the number of ‘‘do not know’’ responses varies by question.
statement that ‘‘everyone enjoys equal rights,’’ whereas the percentage of Tamils
who feel this way is more than three times as high (32%). Only 6% of Sinhala re-
spondents ‘‘strongly disagreed’’ with the statement that ‘‘people are free to speak
their mind without fear,’’ whereas more than four times as many Tamils (25%)
‘‘strongly disagreed’’ with this statement. The percentage of Tamils who responded
that they feel ‘‘less safe’’ than a few years ago is twice as high as among the Sinhala
(41% versus 20%). A surprisingly high percentage of Sinhala (51%) were not sure
that they would be ‘‘treated in the same way as everyone else,’’ but an even higher
percentage of Tamils (73%) were unsure of their equal treatment by the police (see
table 5.1).
Some of our readers may believe that all majority religions, in any country, will
tend toward majoritarianism in politics and public policy. Our SDSA survey
reveals that this is not necessarily so. For whatever complex variety of reasons, the
majority Buddhist population in Sri Lanka, a polity that has followed robust
nation-state policies for fifty years, is much more majoritarian than the majority
Hindu population in India, a polity that has followed state-nation policies for
sixty years.
In Sri Lanka, 89% of the Buddhist majority ‘‘strongly agree’’ or ‘‘agree’’ with the
statement that ‘‘in a democracy the will of the majority must prevail,’’ whereas in
India the comparable response to this question by the Hindu majority is only 48%.
tamils in sri lanka 163
table 5.2
Majoritarianism in Sri Lanka and India by Majority Religion, 2005
Sri Lanka India
(majority Buddhist) (majority Hindu)
Those who ‘‘strongly agree’’ or ‘‘agree’’ 89% 48%
that ‘‘in a democracy the will of the
majority community must prevail.’’
Those who ‘‘strongly disagree’’ or 77 29
‘‘disagree’’ that ‘‘giving equal
treatment is not enough, the
government should give special
treatment to minorities.’’
‘‘Strong majoritarians’’ on composite 27 7
index
Source: SDSA 2005, CSDS, Delhi, Sri Lanka dataset; Sri Lanka–specific questions.
Note: The composite index on majoritarianism uses two items: 31-b (‘‘Giving equal treatment
is not enough, the government should give special treatment to minorities’’) and 31-g
(‘‘Minorities should adopt the ways of life of the majority community’’). ‘‘Strong majoritarians’’
in our index are those who disagree or strongly disagree with 31-b and agree or strongly agree
with 31-g. The overall N for SDSA in Sri Lanka is 4,630. Here the N varies for each question
because the number of ‘‘do not know’’ responses varies by question.
Affirmative action for minorities is much more opposed by Buddhists in Sri Lanka
than it is by Hindus in India. In Sri Lanka, 77% of Buddhist respondents ‘‘strongly
disagree’’ or ‘‘disagree’’ with the statement that ‘‘giving equal treatment is not
enough, the government should give special treatment to minorities.’’ But in
India, only 29% of Hindu respondents ‘‘strongly disagreed’’ or ‘‘disagreed’’ with the
same statement.
The SDSA survey constructed an index of ‘‘strong majoritarianism.’’ Respon-
dents who ‘‘strongly agreed’’ or ‘‘agreed’’ with the statement that ‘‘minorities should
adopt the ways of life of the majority’’ and who also ‘‘strongly disagreed’’ or ‘‘dis-
agreed’’ with the affirmative action question we just discussed in the preceding
paragraph are labeled ‘‘strong majoritarians.’’ In the SDSA survey, only 7% of
Hindu respondents in India are strong majoritarians, whereas four times as many
Buddhists in Sri Lanka (27%) are strong majoritarians. For a summary of these
results, see table 5.2.
As early as the 1970s, Sri Lankan administrations became so aware of the
impending crisis with the Tamils that they began to relax some of the language
strictures against Tamil, but these efforts were never fully implemented, nor the
damage undone. Since the 1990s there have been various peace efforts, but the
164 crafting state-nations
extreme nation-state policies that they had implemented have massively eroded
the Sri Lankan state’s internal capacity even to communicate to their Tamil-
speaking citizens. For example, as late as 1998, the ‘‘National Integration Planning
Unit’’ working with the president on peace plans and quasi-federal proposals sent
out virtually all of its mass mailings to Tamils in Sinhalese. They did this because
they had no Tamil-script computers or Tamil-literate typists in the National Inte-
gration Planning Unit.54
In early 2002, in response to a Norwegian-led effort to mediate a peace process
in Sri Lanka, the Tamil LTTE separatists and the Sri Lankan government agreed
to a ceasefire. The ceasefire was in existence at the time of the 2005 survey but
broke down in January 2008.
In the SDSA survey, a strong majority of Sinhalas and an even stronger majority
of Tamils supported various measures to seek peace. In answer to the statement ‘‘it
is very important to have a peaceful settlement of ethnic conflict for the existence of
a strong democracy in Sri Lanka,’’ 68% of the Sinhalas ‘‘strongly agreed’’ or
‘‘agreed,’’ as did 83% of the Tamils, 94% of the Up-Country Tamils, and 87% of the
Muslims. A comparable consensus supported the statement ‘‘the best way to
achieve peace in Sri Lanka is a solution through negotiation’’: 63% of Sinhalas,
84% of Tamils, 85% of Up-Country Tamils, and 74% of Muslims. Most surprising of
all, almost fifty years after the militant nation-state ‘‘Sinhalese Only’’ monolingual
campaign had contributed to growing polarization and conflict in multinational
and multilingual Sri Lanka, there was nonetheless a strong consensus among all
respondents about the desirability of strong language reforms in a more state-
nation direction: 60% of Sinhalas, 68% of Tamils, 72% of Up-Country Tamils, and
88% of Muslims strongly agreed that ‘‘the state should make it compulsory to learn
all the three languages [Sinhala, Tamil, and English] at school.’’
However, any advance on negotiations during the ceasefire period would al-
most certainly have had to entail constitutionally embedded concessions granting
a serious degree of regional autonomy. On this key issue (after fifty years of unitary
state/nation-state rhetoric from Sinhala-led governments), Sinhala opinion was in
54. The above is based on long interviews (and observations of working arrangements) by
Stepan with Foreign Minister Laksman Kadhirgamar and Minister of Justice and Constitutional
Affairs G. L. Peiris, April 2–6, 1998, Colombo. Kadhirgamar, the only significant Tamil minister
in the Sri Lankan government, was assassinated in 2004, most analysts believe by the LTTE.
In a follow-up interview with a major Sri Lankan political scientist, Jayadeva Uyangoda, in
2004, Uyangoda confirmed Stepan’s impression, saying that while some leaders in the government
were exploring ways for Sinhala-Tamil rapprochement, the state apparatus had still not become
‘‘decommunalized.’’
tamils in sri lanka 165
55. In 1987, the thirteenth amendment to the Constitution was passed, introducing a degree of
self-government at the provincial level. This came in the form of ‘‘Provincial Councils.’’ However, to
date, serious power has never devolved to the Councils. The central government appoints a gover-
nor to the province, and this governor controls the most important part of the budget in the province
and de facto controls education, land, police, and health.
56. Previously cited interview by Stepan with Laksman Kadirgamer. Indeed, the U.S. State Depart-
ment found that staffing in the Sri Lankan armed forces had increased 650% since 1985—but that not
one Tamil had been employed. This situation contrasts sharply with the end of the insurgency in the
1990s in Punjab State in India, where the top police officer during the counterinsurgency was K. P. S.
Gill, himself a Sikh. For an account of how and why Gill led the military campaign against Sikh sepa-
ratists, see K. P. S. Gill, Punjab, The Knights of Falsehood (New Delhi: Har-Anand Publications, 1997).
57. State of Democracy in South Asia project, 2005, Sri Lanka dataset, questions C-13e, C-13f.
166 crafting state-nations
table 5.3
Opinions on Measures to End the Present Conflict by Ethnicity in Sri Lanka, 2005
Up-country
Those who: Sinhalas Tamils Tamils Muslims
‘‘Strongly disagree’’ or ‘‘disagree’’ with 82% 34% 46% 47%
the proposal that ‘‘the powers of some
Provincial Councils may need to be
increased more than others.’’
‘‘Strongly disagree’’ or ‘‘disagree’’ with 66 18 17 24
the proposal that ‘‘the powers of
Provincial Councils should be
increased, even if those of the
government at the center have to be
decreased.’’
‘‘Strongly disagree’’ or ‘‘disagree’’ with 77 31 80 45
proposal for general amnesty.
Favor ‘‘military solution by defeating the 25 3 2 16
LTTE by itself or with negotiations as
the best way to achieve peace in Sri
Lanka.’’
Source: SDSA 2005, CSDS, Delhi, Sri Lanka dataset.
Note: The overall N for SDSA in Sri Lanka is 4,613. Here the N varies for each question
because the number of ‘‘do not know’’ responses varies by question.
Questions:
S-9: ‘‘Now I am going to ask you about some proposals with regarding to the ethnic conflict
that have been discussed recently. For a peace agreement, please state whether you very much
agree, agree, somewhat agree, do not agree with the following proposals:’’
S-9b: ‘‘The powers of some Provincial Councils may need to be increased more than others.’’
S-9a: ‘‘The powers of Provincial Councils should be increased, even if those of the
government at the center have to be decreased.’’
S-9d: ‘‘There should be a general amnesty (that is, freedom from criminal prosecution) for
people who may have committed illegal political violence against civilians during the war.’’
In the last two chapters we have addressed a puzzle: after Independence, why did
the issue of Tamil separatism become a non-issue in India while the non-issue of
Tamil violence and separatism become a war for Tamil independence in Ceylon?
Let us look at comparative levels of trust and distrust in the two countries. In
the state-nation of India, there are surprisingly small differences between majority
Hindu and minority Muslim opinions on such key matters as trust in the army and
perceptions of inequality in the polity. In strong contrast, in Sri Lanka, which has
tried to impose aggressive nation-state policies for fifty years, there are extremely
tamils in sri lanka 167
table 5.4
Opinions on Trust in the Army and Unequal Treatment in Sri Lanka and India, 2005
Sri Lanka India
Majority Minority Majority Minority
Buddhist Hindu Hindu Muslim
Those who have ‘‘no trust at all’’ in the 4% 34% 5% 7%
army
Those who strongly disagree that 9 28 9 8
‘‘everyone enjoys equal rights’’
Source: SDSA 2005, CSDS, Delhi, Sri Lanka dataset.
Note: The overall N for SDSA in Sri Lanka is 4,613. Here the N varies for each question
because the number of ‘‘do not know’’ responses varies by question.
40
35
36
Net Support for Democracy
30
29
25
20 18
15 12
15
10 7
5
0
Low Medium High
Levels of Religiosity
Figure 5.1. Opposite Relationship between Intensity of Religious Practice and Support for
Democracy by Majority Religion in India and Sri Lanka
Notes: Table entries are for ‘‘net support for democracy,’’ which is measured here by
subtracting the percentage of ‘‘non-democrats’’ from the percentage of ‘‘strong democrats.’’
We classify respondents as ‘‘strong democrats’’ if they (1) ‘‘prefer democracy’’ in C23 and
(2) ‘‘strongly approve’’ or ‘‘approve’’ of rule by ‘‘elected leaders’’ in C18d and (3) ‘‘strongly
disagree’’ or ‘‘disagree’’ with rule by army and (4) ‘‘strongly disagree’’ or ‘‘disagree’’ with
rule by the king. We classify respondents as ‘‘non-democrats’’ if they (1) say they prefer
dictatorship or say it doesn’t matter in C23 and (2) either ‘‘strongly disagree’’ or ‘‘disagree’’
with rule by elected leaders or ‘‘strongly agree’’ or ‘‘agree’’ with rule by army or king.
Source: SDSA 2005, CSDS, Delhi, India and Sri Lanka main dataset.
Sri Lanka’s civil war ended in May 2009 with the death of the LTTE leader
Velupillai Prabhakaran and the military victory of the Sinhalese nationalist leader,
President Rajapaksa. The consequences of Sri Lanka’s long conflict over nation-
state grammar in a robustly multinational context had taken a secular toll on
democracy. On Freedom House’s seven-point scale on ‘‘political rights,’’ where
the best score is 1, Sri Lanka never scored worse than 2 from 1973 to 1981; however,
Sri Lanka has not scored as high as 2 since 1983. Indeed, in every full year of
Rajapaksa’s government from 2006 to 2009, Sri Lanka never received a ‘‘political
59. See ‘‘Sri Lanka Notifies Broker Norway of End to Rebel Peace Deal, Nordic Monitors to
Leave,’’ Associated Press, January 4, 2008. There is a rich literature on previously failed devolution
and peace discussions. See, for example, M. Somasundram, ed., Reimagining Sri Lanka: Northern
Ireland Insights (Colombo, Sri Lanka: International Centre for Ethnic Studies, 1999). See also Regi
Siriwardena, ed., Sri Lanka: The Devolution Debate (Colombo, Sri Lanka: International Centre for
Ethnic Studies, 1996).
60. On the vital role of timing in politics, see Juan J. Linz, ‘‘Democracy’s Time Constraints,’’
International Political Science Review 19, no. 1 (1998), pp. 19–37; and the chapter by Linz, ‘‘Time
and Regime Change,’’ in his Robert Michels, Political Sociology, and the Future of Democracy (New
Brunswick, NJ: Transaction Books, 2006). Houchang Chehabi edited this book and added a valu-
able bibliography.
170 crafting state-nations
table 5.5
Contrasting Policies of India and Sri Lanka toward Their Tamil Minority Populations
Policy grammar Tamils in India Tamils in Sri Lanka
1. An asym- The constituent assembly created an No constituent assembly was held, but
metrically federal asymmetrical federal system that enabled the parliament approved a constitution
but not a unitary state boundaries to be redrawn and even- that declared Sri Lanka a unitary state.
state or sym- tually regional cultural majorities to rule After 1956, the Sinhalese-Buddhist ma-
metrically federal these states in their own languages. Non- jority increasingly advanced majoritarian
state Tamil speaking areas were carved out of state policies. No significant devolution-
the then Madras state. ary policies were ever implemented.
2. Individual rights Language: In 1965, after intense mobili- Language: In 1956, Sinhalese was made
and collective recog- zations and political negotiations, plans the only official language.
nition for implementing Hindi as the official
language of the Indian union were aban-
doned. The ‘‘three-language formula’’ for
education was adopted. Tamil is the offi-
cial language of Tamil Nadu, and the
state is not obliged to use Hindi in its
communication with the Union.
Religion: All major religions were consti- Religion: Article 9 of the 1978 constitu-
tutionally recognized. Minority institu- tion assigned Buddhism ‘‘the foremost
tions are eligible for financial support place’’ among religions. State subsidies
from the state. favor Buddhists.
rights’’ score higher than 4. Civil liberties, such as freedom of the press, fared even
worse. Reporters Sans Frontières, in the Press Freedom Index: 2008, noted that for
the first time ever they included Sri Lanka, ‘‘which has an elected government and
where the press faces violence that is all too often organized by the state’’ as one of
the ten worst offenders of the more than 150 countries included in their rankings.61
We stated previously that between 1983 and 2006, Sri Lanka had more estimated
61. See the Asia section of their annual report, available at www.rsf.org/en-classement794-
2008.html. For articles discussing assassinations and imprisonment of journalists, including some of
the most important Tamils, see ‘‘Chronicle of a Death Foretold,’’ The Economist, December 1, 2009,
and the UNHCR, Refworld, September 4, 2009. One of the most prestigious Tamil journalists, three
months after the military triumph, was sentenced to twenty years of hard labor for his reports
considered too sympathetic to the LTTE.
tamils in sri lanka 171
5. Politically inte- Tamils became politically integrated ‘‘Indian Tamils’’ were disen-
grated but not into the Indian polity but main- franchised in 1948; all members of
culturally assimi- tained very strong pride in Tamil pro-autonomist Tamil parties had to
lated populations culture. Different governments in leave parliament in 1983. By the
Tamil Nadu have aggressively taken mid-1980s Sri Lanka was ‘‘robustly
up the promotion of Tamil language politically multinational’’ with polit-
and culture, including in state ically unintegratable Tamil LTTE
schools and educational curricula. guerrilla leaders.
64. For estimates of deaths and casualities in the end-game of 2007–2009, see Michael Renner,
‘‘The Casualties of Sri Lanka’s Intensifying War,’’ World Watch Institute, July 18, 2007; ‘‘War on the
Displaced: Sri Lankan Army and LTTE Abuses against Civilians in the Vanni,’’ Human Rights
Watch, February 2009; and U.S. Department of State, ‘‘Report to Congress on Incidents during the
Recent Conflict in Sri Lanka,’’ available at www.state.gov/documents/organizations/.
65. See the three-part interview with N. Ram, editor-in-chief of The Hindu, July 6, 7, and 8,
2009. These quotations are from the July 6 interview.
chapter six
Ukraine
State-Nation Policies in a Unitary State
Thus far in this book, we have presented nation-states and state-nations as sharply
contrasting ideal types for how to nurture democratic political communities. We
have also noted that a characteristic political institution of the state-nation ideal
type is federalism, indeed asymmetrical federalism. However, theoretically and
empirically, we can imagine geopolitical and domestic contexts where neither full
state-nation nor full nation-state policies are plausible as a way of managing the
multinational dimensions of a polity. Why?
Geopolitically, states are members of a world of states, and it is possible that in
some international contexts, especially if a country has borders with a more mili-
tarily powerful state that has some irredentist tendencies toward it, asymmetrical
federalism, indeed any type of federalism, may present dangers for the nurturing
of a new democratic political community via this classic state-nation policy. The
safest solution might be a unitary state.
Domestically, meanwhile, if politically significant elites are divided deeply
over cultural policies, it would also be democratically dangerous and politically
unfeasible to try to impose classic strong nation-state policies. If we take as given
in some countries such geopolitical and domestic constraints, some key questions
automatically arise. Are mixed state-nation and nation-state policies possible?
Specifically, is it possible to follow many state-nation policies in a unitary state?
Also, is it possible to utilize some ‘‘soft’’ nation-state policies within an overall
state-nation strategy, for example by using some consociational practices without
institutionalizing a full consociational regime?
The goal of our book has not been to extol state-nations over nation-states but
174 crafting state-nations
rather to expand our collective political imaginations about what is feasible, and
unfeasible, in different contexts. One possible context is that which we have just
discussed in which the utilization of the full set of either state-nation or nation-
state policies would seem inadvisable. It is thus incumbent on us in the remainder
of this book to identify and empirically examine some examples of what might be
plausible in such contexts. In this chapter we will explore what political formulas
are possible where aggressive nation-state policies are unlikely to lead toward
societal peace and inclusive democracy but a full state-nation strategy, especially
the use of the key state-nation policy—federalism—would also be dangerous. The
analysis of independent Ukraine will allow us to study such a situation.
We examine Ukraine here from the perspective of three issues. First, we exam-
ine some conditions and identities that make the implementation of aggressive
nation-state policies in Ukraine difficult and dangerous, if the goal is social peace
and democracy.1 Ukraine at independence in 1991 was a multinational society but
not a ‘‘politically robust multinational’’ polity, as we have been using that term in
this book. The data contained in this chapter suggest that if aggressive nation-state
policies had been pursued in Ukraine at independence, self-identifying Russians
in Crimea, and especially some of their political leaders, would most likely (as in
Sri Lanka) have become ‘‘robustly multinational.’’ Some Russians in Crimea,
especially if they felt the only language permissible in the state would be Ukra-
nian, could have become secessionist and, as in Transnistria in Moldova, could
have appealed for and received Russian military support. Second, we will examine
compromises, conditions, and identities that facilitated the use of some state-
nation policies for both the ethnic Russian and the ethnic Ukrainian commu-
nities. Third, we suggest some potential reforms, ones not yet tried in Ukraine,
that might do what many theorists of nationalism think an impossibility: simulta-
neously deepen nation-state as well as state-nation democratic loyalties and iden-
1. Concerning the status of democracy in Ukraine before the Orange Revolution, Lucan A. Way
nicely analyses the formal, and especially informal, ‘‘limitations on democracy [but also] the reasons
why authoritarianism never became consolidated in Ukraine.’’ He asserts that ‘‘Ukraine under
Kuchma’s presidency was a model case of ‘competitive authoritarianism’—a civilian nondemocratic
regime with regularly held elections that are competitive but extremely unfair.’’ See his ‘‘Ukraine’s
Orange Revolution: Kuchma’s Failed Authoritarianism,’’ Journal of Democracy 16 (April 2005),
p. 131. It is probably also useful to note that on Freedom House’s seven-point scale of political rights
(1 being the best score), Ukraine is the only one of the twelve non-Baltic former members of the
Soviet Union never to have had a political rights score in 1992–2005 worse than 4. To be sure, it also
never had a score better than 3. Also, the constitution was deeply debated and was not designed by
the president only, as Russia’s was in 1992, and some features of this constitution were useful to the
Orange Revolution activists. On constitution building in Ukraine, see Kataryna Wolczuk, The
Moulding of Ukraine: The Constitutional Politics of State Formation (Budapest: Central European
University Press, 2001).
ukraine 175
tities without becoming a federal state. In a brief conclusion, we will offer some
comparative reflections concerning whether the unusual amalgam of policies
implied in the title of this chapter is making an inclusive and peaceful democracy
possible in independent Ukraine.
For a newly independent country to opt for aggressive but socially peaceful and
democratic nation-state strategies, it is very useful if three conditions are present:
(1) a strong overlap of the cultural demos and the political polis; (2) a relatively
unified, electorally based political elite that is in agreement about pursuing such
policies; and (3) a geopolitical situation that supports, or at least is not outright
hostile to, the pursuit of such strategies. At independence, Ukraine met none of
these criteria.
1. Cultural demoi, not demos. In a 2005 public opinion survey in Ukraine, 81%
of the respondents self-identified as being of ‘‘Ukrainian nationality’’ and only 17%
self-identified as being of ‘‘Russian nationality.’’2 But identities and everyday lan-
guage use in Ukraine create many more obstacles for aggressive nation-state-
building strategies than these figures indicate. In 1995, for example, in the five east-
ern provinces close to Russia, 59% of the population defined themselves as ‘‘ethnic
Ukrainian’’ but only 15% spoke Ukrainian as their ‘‘language of preference.’’3
Ukraine’s history has created an especially complex multinational society.4
The territory that is now independent Ukraine only attained its present borders in
1954, when Soviet leader Nikita Khrushchev assigned Crimea, the home of the
Soviet Black Sea Fleet, to Ukraine. The majority of the population in Crimea is
made up of ethnic Russians, and there is a sizeable Islamic Tatar population. As
late as 1994, only 6% of the total Crimean population completely supported
Ukrainian independence because they wanted their own sovereignty if not neces-
sarily full independence.5 Important cities in western Ukraine, such as Lviv, until
2. Richard Rose, ‘‘Divisions within Ukraine: A Post-Election Opinion Survey,’’ Studies in Public
Policy, no. 403 (2005), question C-2a, p. 23.
3. See Dominique Arel, ‘‘Ukraine: The Temptation of the Nationalizing State,’’ in Political
Culture and Civil Society in Russia and the New States of Eurasia, ed. Vladimir Tismaneanu
(Armonk, NY: M. E. Sharpe, 1995), p. 170, table 6.5. In all of his work, Arel draws the important
distinction between ‘‘native language’’ and ‘‘language of preference.’’
4. Numerous footnotes and analysis to follow on all these points.
5. See Roman Solchanyk, ‘‘The Post-Soviet Transition in Ukraine: Prospects for Stability,’’ in
Contemporary Ukraine: Dynamics of Post-Soviet Transformation, ed. Taras Kuzio (Armonk, NY:
M. E. Sharpe, 1998), pp. 30–33.
176 crafting state-nations
their annexation by the U.S.S.R. at the end of the Second World War, had for
hundreds of years been a part of the Polish-Lithuanian Commonwealth, the
Austro-Hungarian Empire, and later Poland, while eastern Ukraine was under
Imperial Russia or the U.S.S.R.
Strong nation-state strategy advocates would have had to put great emphasis on
making the Ukrainian language the de facto, as well as de jure, dominant lan-
guage throughout the territory. This would have been difficult because, even if we
exclude Crimea, only 44% of Ukraine’s population shortly after independence
used Ukrainian as their ‘‘language of preference.’’ This percentage varied from a
high of 92% in the west, to only 24% in the capital of the country, Kiev, to less than
15% in the east and south.6
Religion is not a homogenizing factor or even a cross-cutting cleavage in
creating a common cultural demos. Indeed, in many parts of Ukraine, religion
contributes to compounding cleavages. Since the Union of Brest in 1596, many
Orthodox bishops in the West, virtually alone among the Orthodoxy in the world,
broke away from Orthodoxy and pledged allegiance to the Pope in Rome as
Uniates, in return for which they were allowed to retain their Eastern Rites. The
Orthodoxy who are not allegiant to Rome are probably more fragmented (or
pluralistic) in Ukraine than in any country in the world, with three intensely com-
peting Orthodox organizations. Some bishops accept the Moscow Patriarchate,
while others accept the Kyivan (Kiev) Patriarchate. There is also the Ukrainian
Autocephalous Orthodox Church. Finally, and very importantly, a third of the
population self-identify as ‘‘not religious.’’7
From August 10 to September 29, 1991, Richard Rose, in collaboration with the
All-Russian Public Opinion Research Center (VCIOM), conducted a survey
among the Russian minorities in eleven of the fifteen republics of the U.S.S.R.8
The responses from self-identifying ethnic Russians in Ukraine graphically docu-
ment how, on some critical attitudes, only a few months before independence,
they were very far from feeling like members of the Ukrainian nation-state, or
6. Arel, ‘‘Ukraine: The Temptation of the Nationalizing State,’’ in Political Culture and Civil
Society in Russia and the New States of Eurasia, ed. Tismaneanu, p. 170.
7. See José Casanova, ‘‘Ethno-Linguistic and Religious Pluralism and Democratic Construc-
tion in Ukraine,’’ in Post-Soviet Political Order: Conflict and State Building, ed. Barnett R. Rubin
and Jack Snyder (London: Routledge, 1998), pp. 81–103, esp. table 5.2. In the 2004 presidential
election, followers of the Kyivan and Autocephaelous Orthodox churches tended to support the
Orange Revolution, whereas priests and monks from the Moscow Patriarch often paraded in support
of Moscow-supported Yanukovych. ‘‘Ukrainians Threaten Orthodox Split’’ in The Ukraine List,
no. 355, pp. 17–20.
8. For the Gorbachev and early post-Gorbachev period, VCIOM was by far the best Moscow-
based survey institute for Russia and key post-Soviet states like Ukraine.
ukraine 177
‘‘demos.’’ In answer to the question, ‘‘With what nationality do you define your-
self ?’’ 89% answered ‘‘Russian’’ and only 7% answered ‘‘Ukrainian.’’ In answer to
the question, ‘‘When you say ‘in our country’ what do you most often mean?’’ 79%
answered ‘‘Soviet Union,’’ 11% answered ‘‘this republic.’’ Concerning the inquiry
as to the language spoken by their children and grandchildren, 67% said it was
Russian, 2% said it was Ukrainian, and 29% said it was both. Significantly, 52% of
the ethnic Russians felt that knowledge of the Russian language should be com-
pulsory for ethnic Ukrainians. In an important political sense, the ethnic Russians
were happy if the U.S.S.R. or the new state of Russia retained some political
responsibility for protecting the Russian diaspora in Ukraine. In answer to the
question, ‘‘Who should be responsible for protecting Russians’ interest in the
republic?’’ (with multiple answers allowed), Russians in eastern Ukraine indi-
cated Republican leadership (74%), U.S.S.R. leadership (44%), Russian leader-
ship (40%), and the Soviet Army (7%).9
2. Existing sharp divisions among the major politically relevant elites. National-
ism and the demand for independence had an early champion in the People’s
Movement of Ukraine, also called Rukh, which later became a political party.
Among these early nationalists, there were certainly many who would have liked
to press for strong nation-state policies. However, unlike Poland’s Solidarity, Rukh
had trouble building a strong and active constituency outside of its original base,
which was in the predominantly Ukrainian-speaking, pro-European, anticommu-
nist, western part of Ukraine centered around Lviv (with some strong followers in
the capital Kiev in the center of the country). Rukh was close to the zenith of its
powers in 1989, but at its First Congress 85% of its members were from the west
and the center and only 6% from eastern Ukraine.10 In the 1991 presidential
elections, the Rukh candidate came in second, with 23% of the vote. In the first
two competitive elections for the parliament, in 1990 and 1994, Rukh and its allies
managed to win only slightly more than a quarter of the seats.11
The other major political elite group came from the Ukrainian Communist
Party (CPU). The former ideological secretary of the Communist Party, Leonid
Kravchuk, won the first presidential elections in 1991 with 62% of the vote. In the
9. See VCIOM (with Richard Rose), ‘‘Russians outside Russia: A 1991 Survey’’ and ‘‘Russians in
the Baltic: A 1991 Survey,’’ Studies in Public Policy, nos. 283 and 287 (1991), respectively, Center for
the Study of Public Policy, University of Strathclyde, Glasgow, 1997.
10. Wolczuk, Moulding of Ukraine, p. 96.
11. On the 1990 elections, see Arel, ‘‘Parliamentary Blocs in the Ukrainian Supreme Soviet,’’
Journal of Soviet Nationalities 1, no. 4 (Winter 1991), pp. 108–154. On the limits these elections
presented to any ethnic-based nationalism, see Roman Szporluk, ‘‘Reflections on Ukraine after
1994,’’ Harriman Review 7 (March–May 1994), p. 2.
178 crafting state-nations
first parliamentary elections of 1990, the CPU emerged as the largest party.12 In
fact, while many of the 450 members of the Ukrainian parliament in 1990 were
formally listed as independents, Wolczuk asserts that ‘‘only 68 deputies were not
CPU members,’’ although real partisan membership was certainly less strong than
these data indicate because, among other things, an individual could be a mem-
ber of CPU and Rukh.13 With a strong base among Russian-speakers, especially in
the eastern parts of Ukraine, the Communist Party and its allies had very few
incentives to champion aggressive nation-state policies that would lead to a mar-
ginalization of the Russian language.
There were in fact some disincentives. A rapid movement toward Ukrainian as
the only usable language in parliament may well have hurt the careers of the
members of the largest party in the parliament, the CPU. A content analysis of
speeches in the parliament reveals that only 30% of CPU deputies spoke Ukrai-
nian in parliamentary debates in the early 1990s.14
3. Geopolitical constraints on aggressive nation-state policies. At the time of
Ukranian independence, it was by no means clear how strong irredentism might,
or might not, become in the rump state of Russia. However, what was clear is that
by far the largest number of ethnic Russians who were ‘‘lost to Russia’’—11.3
million—were in Ukraine.15 Also, Ukraine still had a much greater density of
Russian soldiers on its territory than did Russia itself. At independence, Ukraine
had a soldier-to-inhabitant ratio of one soldier for every 98 inhabitants, whereas
Russia only had one soldier for every 634 inhabitants—and many of these Russian
soldiers in Ukraine were still under Russian commanders supported by a still
world-class military system.16 Moreover, prominent Russian citizens from elite
institutions repeatedly referred to Ukrainian independence as ‘‘temporary.’’ Nu-
merous public opinion polls by the Moscow-based Public Opinion Foundation
found that ‘‘Russians could not accept Ukrainian independence.’’17
In comparative terms, it is important to stress that geopolitically, none of
the multinational societies that we consider close to the state-nation pole, such
as India, Belgium, Spain, and Canada, had at independence a potentially hos-
tile irredentist relationship with a militarily more powerful neighboring country.
Ukraine, at independence, potentially had such a politically significant irredentist
relationship with Russia.
To some extent Ukraine has a mutually dependent energy situation, because
most of Russia’ exports to Europe pass through Ukraine; however, in a short-term
crisis, Ukraine is extremely vulnerable to Russian retaliation. Estimates are that as
much as 90% of Ukrainian oil and 60% of Ukrainian gas comes from Russia, at
highly subsidized prices.18 On a few occasions, Russia has literally turned the
lights out in parts of Ukraine.19 Geopolitical as well as domestic power balances
therefore made aggressive nation-state policies in the aftermath of independence
both difficult and dangerous.
Two months after the Orange Revolution, Stepan asked one of the world’s leading
political science specialists on Ukraine, Dominique Arel, the following question:
‘‘If, at independence, a full nineteenth-century French-style nation-state strategy
had been attempted, what are the chances that Ukraine would have remained
united and peaceful?’’ Arel’s answer: ‘‘Nil.’’20
There was certainly no conceptually conscious effort to follow strategies close
to what we have called the state-nation ideal type, but there were conscious ef-
forts to avoid the nation-state dangers.21 Stepan first became aware of this in the
18. See the discussion by Sherr, ‘‘Ukrainian Security Policy,’’ p. 263. This situation led some
wags to say that ‘‘in the spring Ukraine leans toward the West, but in the fall toward the East.’’
19. Communication with an IDB official based on his personal observations in Eastern Ukraine,
London, June 2005. See Jonathan Stern, ‘‘The Russian-Ukrainian Gas Crisis of January 2006,’’ Oxford
Institute for Energy Studies, available at www.oxfordenergy.org/pdfs/commente0106.pdf. See also
‘‘Russia, Ukraine Gas Dispute Turns Ugly,’’ Associated Press, December 14, 2005.
20. Dominique Arel, chair of Ukrainian Studies at the University of Ottawa and coordinator of
the Ukraine List newsletter, is the author of numerous works on Ukraine and comparative national-
ism. He coordinated the Ukraine section of the comprehensive project on ‘‘Identity Formation’’ led
by David D. Laitin. Stepan’s interview with Arel was in February 2005 at the University of Ottawa.
We should note that a standard assertion in the literature on Ukraine is that there was virtually no
ethnic or political violence during or after the drive for independence. In 1994, there were a rash of
articles about the possible breakup of Ukraine. Thus the formulation of Stepan’s question.
21. However, José Casanova cites an early discussion by Juan Linz of our state-nation concept
and argues that it was the most appropriate strategy for Ukraine to follow and that many of the poli-
cies actually negotiated, and implemented, were close to state-nation policies. Casanova, ‘‘Ethno-
Linguistic and Religious Pluralism,’’ esp. pp. 87–88. Similarly, Roman Szporluk does not use the
state-nation terminology, but the implications are consistent with our argument. Szporluk, ‘‘Reflec-
tions on Ukraine after 1994.’’
180 crafting state-nations
22. For a well-documented article that argues that Ukranian nationalists appreciated this multi-
national power diffusion reality and looked for ways to accommodate it inside newly independent
Ukraine, see Paul D’Anieri, ‘‘Ethnic Tensions and State Strategies: Understanding the Survival of
the Ukrainian State,’’ Journal of Communist Studies and Transition Politics 23 (March 2007), pp. 4–
29. This article contains bibliographic references to a CIA report discussing the very real threat of
the breakup of Ukraine by the CIA. Journals such as The Economist, Foreign Policy, and Forbes
magazine also contained similar reports.
23. Much documentation supportive of the reality of multiple but complementary identities in
Ukraine is found in Henry E. Hale, Foundations of Ethnic Politics (Cambridge: Cambridge Univer-
sity Press, 2008).
24. A separate sample was made for Russians in western Ukraine.
ukraine 181
§ ‘‘All people in the republic, both Russians and titular residents, share the
same hardships now.’’ (Agree: 96%)
§ ‘‘Russians in this republic are being discriminated against.’’ (Agree: 3%)
§ ‘‘I don’t feel alien in this republic.’’ (Agree: 86%)
§ ‘‘How would you evaluate relations between different nationalities in the
republic?’’ (Cordial: 28%; Normal: 58%; Tense: 9%; Hostile: 0%)
§ ‘‘Do you intend to emigrate from this republic or do you intend to stay for
the rest of your life?’’ (Firmly decided to emigrate: 2%)
§ ‘‘Pogroms of Russians are possible in the foreseeable future.’’ (Agree: 3%)25
25. VCIOM (with Rose), ‘‘Russians outside Russia,’’ questions 43, 44, 41, 81, 31, and 48, respec-
tively. In Latvia and Estonia, the titular populations were attempting nation-state policies, and
Russian-minority answers on these six questions were much more wary. The only area where
Ukrainian nation-state advocates were in the majority at the time of the survey was in western
Ukraine, and Russian-minority attitudes in western Ukraine were substantially more skeptical of the
neutrality and fairness of a potential Ukrainian state.
26. ‘‘Reflections On Ukraine in 1994,’’ p. 1.
27. Wolczuk, Moulding of Ukraine, p. 89.
182 crafting state-nations
eastern part of the country. The compromise was also deliberately crafted as a way
of blunting the warnings by Communist Party hardliners on the dangers of ethnic
exclusion modeled on the Baltics.28
4. Growing strategic cooperation between Rukh and the communist leadership
over statehood. As the Rukh-based nationalists began to realize that their elec-
toral upper base was probably only between 20% to 30% of the vote and that
they needed the cooperation of Communists to win and sustain an indepen-
dent Ukrainian state, they increasingly shifted from an exclusive, ‘‘nation-based’’
discourse to a more inclusive, ‘‘state-based’’ one. As Alex Motyl and Bohdan
Krawchenko note, ‘‘Rukh’s language was palatable to the Communists such as
Kravchuk because it was nationalist, but neither chauvinist, nor racist; it had at
its core the attainment of statehood for the Ukrainian people, whom Rukh care-
fully defined in non-ethnic terms that permitted Russians, Jews, and Poles, and
others to take part in and support its cause. Such a nationalism was at least
as potentially appealing to Communists, as it promised them the opportunity
of continuing to serve as an elite, if not the only elite, within a future Ukrai-
nian state.’’29
With the two major political elite groupings cooperating strategically in their
advocacy of independence and an inclusive citizenship law already passed, on
December 1, 1991, a republic-wide referendum approved the independence of
Ukraine by 90%, including 80% support in the heavily Russophone eastern oblasts
(regions) and 54% in Crimea.30 On the same day, the once anti-nationalist but
now pro-independence Kravchuk was elected president of Ukraine with 62% of
the vote.
5. Partial compromises over the constitutional preamble. As in many countries
with a multinational society, there was a heated constitutional debate over the
preamble. Who are the people? The preamble to the constitution adopted in 1996
was indeed a partial victory for the nationalists, but in the same long and tortuous
sentence, there is a partial compromise for all nationalities. The parliament voted
for the constitution ‘‘on behalf of the Ukrainian people—citizens of all nationali-
ties, expressing the sovereign will of the people, based on the centuries-old history
28. Ibid.
29. Motyl and Krawchenko, ‘‘Ukraine,’’ pp. 250–251. Casanova is in agreement. He writes that
‘‘Rukh had incorporated explicitly such an inclusive policy in its platform.’’ Casanova, ‘‘Ethno-
Linguistic and Religious Pluralism,’’ p. 86. The fact remains, however, that the core membership
and electorate of Rukh retained a Ukrainian ethnic base.
30. The incentives for strategic cooperation by the communist elites were enhanced because
some of them supported the failed coup of August 1991 in Moscow and wanted to be protected from
reprisals by Yeltsin.
ukraine 183
§ In Ukraine, the free development, use and protection of Russian, and other
languages of national minorities of Ukraine, is guaranteed.
Article 53 further entrenches Russian and makes illegal any aggressive nation-
state, homogenizing, ‘‘Ukrainian Only’’ state policies:
33. Many prominent Tamils did have a second language, but it was English.
34. See chapter 5. Since 1997 Stepan has made three research trips to Sri Lanka. Virtually all the
key participants he interviewed, from the prime minister on down, now think that their aggressive
nation-state policies, especially toward the privileged place of Sinhalese and Buddhism, and their
complete hostility to decentralization contributed directly to the crisis.
ukraine 185
adopted in Sri Lanka, policies that make state-nation identities possible, notwith-
standing the fact that Ukrainian is the only official language of the country. The
pattern of the everyday use of language in Ukraine helps defuse the issue. In a
1995 poll, 32% of respondents said that they used Ukrainian as their predomi-
nant language at home, 33% used Russian as their predominant language, and
35% used both Ukrainian and Russian.35 In any case, in Ukraine, the language
of preference is not necessarily a marker of self-identity. Although only 32% of
the population said that Ukrainian was the predominant language they spoke at
home, 70% of them nonetheless self-identified as ethnic Ukrainians.36 Finally, the
Russian and Ukrainian languages are sufficiently close that in parliament, on
television, and elsewhere, it is common for conversations to proceed with one
person that speaks exclusively Russian while the other person speaks exclusively
Ukrainian, and often speakers switch languages in the same sentence.37 Thus,
whatever language is spoken, there is normally not a problem of complete mutual
incomprehension, at least in the linguistic sense.
Those self-identified ethnic Ukrainians who are in fact Russophone under-
standably might not want to jeopardize their careers by having to speak Ukrainian
at the workplace or to use it in exams for entry into the public service. However,
while only 40% of self-identified Ukrainians in Kiev spoke Russian at home, 98%
of them said that they wanted their children to be ‘‘fluent’’ in Ukrainian.
A promising sign for the long-term possibility of multiple but complementary
identities in a possible state-nation polity in Ukraine is that over 90% of self-
identified ethnic Russians in Lviv expressed a desire that their children be fluent
in Ukrainian.38
7. The use of some key common symbols. If a multinational society is to have a
political community close to the state-nation ideal type, its citizens and political
leaders need to have recourse to some common symbols.39
In Ukraine, there are more facilitating conditions concerning nonpolar identi-
ties than the huge literature on the ‘‘two Ukraines’’ indicates. Fortunately, an
important longitudinal study of the supposedly polar opposite cities, Lviv in the
35. In 2005, the figures were 42%, 36%, and 22% respectively. See Oleh Protsyk, ‘‘Majority-
Minority Relations in the Ukraine,’’ JEMIE 7 (2008), p. 24. JEMIE is a publication of the European
Centre for Minority Issues.
36. Ibid. This is a standard finding in surveys on Ukraine.
37. This is called ‘‘Surzhyk,’’ or ‘‘code switching.’’
38. See Ian Bremmer, ‘‘The Politics of Ethnicity: Russians in a New Ukraine,’’ Europe-Asian
Studies 46, no. 2 (1994), table 12.
39. For example, the king in Spain and Belgium and, as we showed in chapter 4, Nehru and
Gandhi for citizens in Tamil Nadu and India are shared positive symbols throughout the polities.
186 crafting state-nations
so-called nationalist west and Donetsk in the so-called pro-Russian east, has been
carried out by the Ukrainian scholar Yaroslav Hrytsak. In this study, the over-
whelming percentage of people polled agreed that the medieval Kyivan Rus
period was the starting point of Ukraine. There are thus not competing but rather
complementary founding myths about Ukraine. Also, throughout all of Ukraine,
the most popular historical figure is the Cossack leader Bohdan Khmelnytskyi,
whom both ethnic Ukrainians and ethnic Russians in Ukraine (unlike Russians in
Russia) see as a major state-builder. Indeed, even President Leonid Kuchma,
often accused of pro-Russian tendencies, made a major speech on the four hun-
dredth anniversary of Khmelnytskyi’s birth suggesting his ‘‘we-feeling’’ with him:
‘‘Today, carrying on Bohdan’s work, we are realizing the third attempt at the
revival of our state.’’40 In fact, Kuchma’s most important book, which is half-
memoir and half-history, has a title that translates as ‘‘Ukraine Is Not Russia.’’
The theoretical literature on nationalism often points to the utility of an
‘‘other’’ in forging common identities. In Ukraine, ethnic Ukrainians, ethnic
Russians, and Crimean Tatars have a shared historical ‘‘other’’—Stalin. The fam-
ine in Ukraine, which led to millions of deaths, and the expulsion of the Tatars
from Ukraine in 1944, are seen as direct results of the policies of Stalin. Thus,
ethnic Russians and ethnic Ukrainians alike ‘‘both in the West and East, are
unanimous in their negative evaluation of Stalin and his acts of repression; they
see him as the main villain in Ukrainian history, the number one anti-hero. And
this is exactly what makes them different from Russians [in Russia]: a majority of
the Russian population [in Russia] considers Stalin in rather positive terms, as a
great state builder, who turned the Soviet Union into a world superpower.’’41 A
more recent pre-independence appearance of an external ‘‘other’’ that forged
some commonalities among virtually all citizens of Ukraine, vis-à-vis Russia, was
the Chernobyl disaster and its callous handling by Moscow.42
Too many observers conflate voter, identity, and issue polarization. To understand
the possibilities of some state-nation policies in Ukraine, it is imperative that we
disaggregate polarization. ‘‘Voter polarization’’ in Urkraine is of course by com-
parative standards very high. But ‘‘identity polarization’’ is less so, and there is far
less ‘‘issue polarization’’ that most of the literature indicates.
To be sure, as the literature on the ‘‘two Ukraines’’ suggests, there are strong
voting differences between Lviv and Donetsk and what would appear to be dan-
gerous voter polarization between East and West.43 For example, in the third and
final round of voting for the president during the Orange Revolution, the pro-
Orange Viktor Yushchenko won 93.7% in Lviv but only 4.2% in Donetsk.44
But, a few years earlier, when a hard question was asked of the population of
the two cities if they should become divided into several different countries, only
1% of Lviv residents and 5% of Donetsk residents chose the option of radically
dividing the political community.45 Conclusion: voter polarization does not nec-
essarily reflect polarization about desired future political identities.
Fortunately for the political community of Ukraine, voter polarization is only
one type of polarization. Conceptually and empirically, we can also speak (and
measure) identity polarization and issue polarization as well as voter polarization.
Identity polarization is substantially lower in Donestsk than voter polarization.
For example, on a three-point scale of identity, only 20% of respondents said they
were ‘‘only Russian,’’ but 28% said they were ‘‘only Ukranian,’’ and the modal
response was bi-ethnic, at 48%. Thus, 76% of respondents in Donetsk have some
degree of Ukrainian in their self-definition of identity.
43. For an excellent review of the extensive literature on the polarization debate in Ukraine and
a very informative quantitative analysis that leads the authors to the conclusion that it makes more
sense to speak of eight, as opposed to two, regions, see Lowell W. Barrington and Erik S. Herron,
‘‘One Ukraine or Many? Regionalism in Ukraine and its Political Consequences,’’ Nationalities
Papers 32, no. 1 (March 2004), pp. 53–86.
44. See Rose, ‘‘Divisions within Ukraine,’’ pp. 49–50. In the 2004 U.S. presidential elections, the
most polarized state was Utah (Bush 71%, Kerry 26%), but the District of Columbia resembled
Donetsk in that the losing presidential candidate won overwhelmingly (Kerry 90%, Bush 9%). One of
the leading Ukraine analysts, Andrew Wilson, after a July 2005 visit to Donetsk, wrote that polarization
was declining somewhat and that ‘‘most east Ukrainian elites are regrouping under party labels that
accept the agenda set by the new regime.’’ See The Ukraine List, no. 354, July 15, 2005, p. 11.
45. See the fascinating studies by Yaroslav Hrytsak, ‘‘National Identities in Post-Soviet Ukraine:
The Case of Lviv and Donetsk,’’ in Cultures and Nations of Central and Eastern Europe, ed. Zvi
Gitelman, Lubomyr Hajda, John-Paul Himka, and Roman Solchanyk, Harvard Ukrainian Studies,
vol. 22 (1998), pp. 263–282, and ‘‘Ukrainian Nationalism, 1991–2001: Myths and Perceptions,’’ Aus-
trian Institute for Eastern and South-Eastern Europe, Vienna, Austria, October 15, 2001.
188 crafting state-nations
Issue polarization is also strikingly lower than voter polarization. The north-
west, where Lviv is located and which voted overwhelmingly for Yushchenko in
2004, was presented in the Russophone media during the 2004 elections as sup-
porting extreme ‘‘Ukrainian Only’’ language plans and opposing good trade rela-
tions with Russia. In fact, 68% of those surveyed in the northwest, far from being
in favor of more restrictive policies toward the status of Russian were actually in
favor of more permissive policies toward the Russian language than the constitu-
tion prescribed: 45% were in favor of Russian being made an official language at
the local level if people want it, and 23% were in favor of Russian being made the
second official language. In terms of Ukrainian government relations with West-
ern Europe and Russia, there was surprisingly little issue polarization in north-
western Ukraine. On a five-point scale, 56% of respondents in the northwest chose
the middle position, ‘‘equal orientation toward the West and toward Russia.’’ This
was also—barely—the modal position (36%) chosen in the southeast.46
These data offer some hope that, provided that future electoral campaigns are
waged in a context of a freer press and that less government-controlled ‘‘admin-
istrative resources’’ are used to generate polarization than there was in the 2004
elections and that the pro- and anti-Orange polarization of the 2004, 2006, and
2008 struggles dies out, new politicians might find it increasingly productive to
devise policies that appeal to citizens on issues where there is in fact not over-
whelming polarization.47 If so, voter polarization could diminish, and multiple
but complementary identities of the state-nation sort could increase. However, it
should be noted that in the 2010 presidential election, geographical polarization
was still powerful.48
The above heading may strike some readers as a contradiction in terms, because
they may assume that the two ideal types are always in an either/or relationship.
Let us see why this is not necessarily so, either theoretically or empirically.
46. All these data are drawn from Dominique Arel and Valeri Khmelko, ‘‘Regional Divisions in
the 2004 Presidential Elections in Ukraine,’’ paper written for the First Annual Danyliw Seminar on
Contemporary Ukrainian Studies, University of Ottowa, Canada, September 29–October 1, 2005.
47. In a private note, Henry E. Hale commented to Stepan that this may be beginning to
happen in that in the 2010 presidential election, a ‘‘new generation’’ type of candidate, Sergei
Tyhypko who tried to bridge the east/west divide came in a surprisingly strong third.
48. For example, the map of electoral results, which depicted Yanukovych in blue, shows the
east and southeast as a solid blue. See Kyiv Post, January 18, 2010, available at www.kyivpost.com/
news/politics/detail/57376/.
ukraine 189
The goals of both ideal types include the creation of a useable state, without
which a democracy is impossible. Both ideal types can also have policies that may
help develop commitment and loyalty to the democratic institutions of the state.
We have argued that Ukraine has a better chance of creating a democratic politi-
cal community if it does not pursue aggressive nation-state policies. However, in
theory, there can be policies that may look like nation-state policies but that, if
implemented softly and widely, also facilitate the multiple but complementary
identities that are crucial both for state-nations and for democracies in multi-
national societies. Let us explore this proposition in the case of Ukraine.
Adult Russophones do not want to be forced to speak Ukrainian in the work-
place or in the exercise of their full rights of citizenship. Nonetheless, the over-
whelming percentage of them want their children to have the opportunity to
speak Ukrainian fluently. However, in many schools in the east and the south,
there are not enough Ukrainian-speaking teachers, and many of the textbooks are
not only in Russian but were also written in the Soviet era in what is now Russia.49
If Russophone parents see that state policies are improving their children’s capac-
ity to function successfully in Ukraine, these same Russophone parents might well
increase their trust in the state-nation policies of the Ukrainian state. If in two
generations, virtually every adult in Ukraine speaks Ukrainian (even though many
of them will continue to speak Russian), the policy would also have been also
been a good nation-state policy.
A somewhat more complex matter is the role of Russian-language news on
television. Because of financial constraints and sensitivity to nation-state sensibili-
ties, the Ukrainian government television channels, such as UT-1, are in Ukrai-
nian. This meant that most Russophones, especially in eastern Ukraine, watched
Moscow-originated programs during the 2004 presidential elections and during
the Orange Revolution. The production of high-quality television news programs
in Russian by the time of the Orange Revolution may have been seen as a linguis-
tic compromise on nation-state goals, but it would probably would have been a
plus for building a political community in Ukraine that is at ease with Ukraine as a
viable democratic nation-state for some of its citizens and a viable democratic
state-nation for other of its citizens. Certainly the goal of having Russophones with
multiple but complementary identities would have been served.
What about federalism? Every country close to the state-nation pole (with the
exception of tiny Luxembourg) is federal. Must Ukraine become federal to main-
49. See Nancy Popson, ‘‘The Ukrainian History Textbooks: Introducing Children to the ‘Ukrai-
nian Nation,’ ’’ Nationalities Papers 29, no. 2 (2001), pp. 325–50.
190 crafting state-nations
tain itself as a country with many state-nation qualities? Most of the scholars we
have asked are somewhat worried about territorially concentrated power under
the existing conditions of polarization and policy conflicts with its large neighbor
Russia. As we discussed earlier, Ukraine—unlike the near-state-nations of Spain,
India, Belgium, or Canada—had a potentially threatening irredentist situation on
its borders. Under such circumstances, federalism might not have been prudent.
However, is federalism nonetheless the only solution? Not necessarily. But serious
decentralization and center-regional development projects should be considered.
Many of the east-west conflicts find their roots in economic as much as cultural
factors. For example, part of the reason that Viktor Yanukovych received 93.5% of
the vote in Donetsk in the third round of the 2004 presidential election was that
industrial production in Donetsk had declined by nearly 60% in the 1990–98
period. However, Yanukovych was appointed governor of Donetsk in 1997. He
worked with the central government on a series of industrial promotion poli-
cies and created two new ‘‘special economic zones.’’ By 2003, Donetsk’s indus-
trial production had improved substantially, the unemployment rate had declined,
and Yanukovych received much of the credit. Correctly or not, Yushchenko was
presented in the 2004 elections as an opponent of these regional development
zones.50
Much of eastern Ukraine is dominated by state-owned Soviet-era mines and
factories. At the moment, every governor is appointed directly by the president,
and municipal budgets are small. A model for Ukraine to consider is the Scan-
dinavian countries, which combine unitary states with much greater budgets for
cities than the federal states of Europe.51
50. These figures on Donetsk come from Svitlana Kalinina, Alexander Lyakh, Galina Sav-
chenko, and Adam Swain, ‘‘Regional ‘Lock-In’ or Local Hegemonic Bloc? Industrial Restructuring
in the Ukrainian Donbas,’’ paper prepared for the Danyliw Research Seminar in Contemporary
Ukrainian Studies at the Chair of Ukrainian Studies, University of Ottawa, September 29–October
1, 2005. See also Barrington and Herron, ‘‘One Ukraine or Many,’’ pp. 70–71. From a comparative
viewpoint of political symbols, it should be noted that in Donetsk, the Ukrainian flag is flown (unlike
Kurdistan, where the Iraqi flag is not flown), and no one has reported a Russian flag flying in
Donetsk, as some were briefly flown in Crimea.
51. For example, from 1992 to 1996 municipalities in the four democratic federal systems of
Western Europe spent 14.3% of total public expenditures, whereas municipalities in the eleven
unitary states spent 24.6% of total public expenditures. In Denmark, Finland, and Sweden, munici-
pal expenditures in this same period accounted for 44%, 37%, and 35% of total public expenditures,
respectively. See Government Finance Statistics (Washington, DC: International Monetary Fund),
various issues. For an analysis of these robust municipal expenditures in unitary states, see Juan J.
Linz and Alfred Stepan, ‘‘Inequality Inducing and Inequality Reducing Federalism: With Special
Reference to the ‘Classic Outlier’—The USA,’’ paper given at the 18th Congress of the International
Political Science Association, Quebec City, Canada, August 1–5, 2000, esp. table 2.
ukraine 191
What can we say after nearly twenty years of Ukraine’s independence and its use of
many more state-nation policies than nation-state policies? Specifically, what can
we say about violence, stateness, democracy, and the possibility of moving toward
power-sharing arrangements normally associated with parliamentarism?
We have argued that one of the goals of state-nation policies is to help facilitate
domestic peace and a high degree of stateness in multinational polities. There
had been, of course, much speculation in many major journals that Ukraine
might fragment. The Economist, for example, ran an article called, ‘‘Ukraine: The
Birth and Possible Death of a Country,’’ a Foreign Policy article was entitled,
‘‘Letter from Eurasia: Will Ukraine Return to Russia?’’ and the Central Intel-
ligence Agency leaked a report about the likely fragmentation of Ukraine.52
However, in a recent review of ethnic tension and state survival in Ukraine, Paul
D’Anieri wrote that ‘‘secessionist movements have made few inroads, and violence
has been non-existent.’’53 Dominique Arel asserted that ‘‘regionalism in Ukraine does
not constitute a threat to the territorial integrity of Ukraine. . . . Russian-speaking
elites and the general population in the East seek inclusion, not separation.’’54
This peaceful situation in Ukraine contrasts sharply with Sri Lanka, where, as
we documented earlier, aggressive nation-state policies led to civil war and nearly
to state disintegration. Ukraine’s peace is also in sharp contrast to two of its Black
Sea post-Soviet neighbors, Moldova and Georgia, both of which lost sovereignty
over part of their states largely due to conflicts over their hardline nation-state
policies in multinational settings.
In Moldova, the nation-state-building Moldovian Popular Front at indepen-
dence pushed hard for Romanian as the national language, for union with Ro-
mania, and for a radical downgrading of the Russian and Ukrainian languages
52. Full references to all these reports and others are contained in Paul D’Anieri, ‘‘Ethnic
Tensions and State Strategies.’’
53. Ibid., p. 4.
54. Dominique Arel, ‘‘The Hidden Face of the Orange Revolution: Ukraine in Denial Towards
Its Regional Problem,’’ a translation of ‘‘La face cachée de la Révolution Orange: l’Ukraine en
négation face á son problèm régional,’’ Revue d’études comparatives Est-Ouest 37 (December 2006),
p. 38.
192 crafting state-nations
widely spoken in the Transnistria district of Moldova: ‘‘In August and September
1989, the Moldovan parliament adopted a number of interrelated language laws.
Moldovan/Romanian was declared the sole state language; in the future it was to
be written with Latin characters. While the legal texts in themselves were rather
liberal, allowing for substantial use of the minority languages in a transitional
period, nationalist enthusiasts among Moldovan bureaucrats, locally and in the
capital, often gave them a most wanton interpretation and used them as a weapon
of linguistic discrimination.’’55 Lucan Way, a close follower of Moldovan politics,
writes that most analysts ‘‘agree that the ethnic tension and brief civil war/separa-
tion of Transnistria [which endures to this day] was facilitated by pro-Romanian
rhetoric of the front.’’56 From March 2 to July 21, 1992, there was what was called
the ‘‘War of Transnistria.’’ In the final stage of the civil war, the former Soviet 14th
Guards Army entered the conflict, opening fire and deploying combat tanks.
Since then, Moldova has exercised no effective control over Transnistria.57 In
1995, the former Soviet 14th Army Unit in Transnistria was redesignated the
‘‘Operational Group of Russian Forces in Transnistria.’’58 The former Soviet com-
bat unit still consisted of ‘‘one motorized infantry division and a number of spe-
cialized subunits, 6,000 men in all. . . . In the army depots enormous quantities of
military hardware, ammunition, and weaponry remained.’’59
In Georgia, from August to September 2008, the interaction between growing
autonomous efforts in the ethnically and linguistically distinct regions of South
Ossetia and Abkhazia, and Georgia’s aggressive nation-state effort towards them,
led to militarized violence in both areas and to an eventual war between Russia
and Georgia. Russia not only defeated Georgia but also rapidly recognized South
Ossetia and Abkhazia as independent countries.60 In February and April 2010,
Russia signed agreements with Abkhazia and South Ossetia to place military bases
in both territories.61
55. Pȧl Kolstø and Andrei Malgin, ‘‘The Transnistrian Republic: A Case of Politicized Regional-
ism,’’ Nationalities Papers 26, no.1 (1998), p. 107.
56. Personal correspondence between Alfred Stepan and Lucan Way, April 22, 2010. See also
William Crowther, ‘‘The Politics of Ethno-National Mobilization: Nationalism and Reform in
Soviet Moldavia,’’ The Russian Review 50 (April 1991), pp. 183–202.
57. On how strong nation-state demands for the use of Romanian in Moldova contributed
to the involvement of the Russian military, see William Crowther, ‘‘Politics of Ethno-National
Mobilization.’’
58. Andreas Johannson, ‘‘The Transnistrian Conflict after the 2005 Moldovan Parliamentary
Elections,’’ Journal of Communist Studies and Transition Politics 22 (December 2006), p. 509.
59. Kolstø and Malgin, ‘‘Transnistrian Republic,’’ p. 111.
60. On the war in Georgia and its impact in Ukraine, see Dominique Arel, ‘‘Ukraine since the
War in Georgia,’’ Survival 50, no. 6 (December 2008–January 2009), pp. 15–25.
61. See ‘‘Russia Recognizes South Ossetia and Abkhazia to Save People’s Lives,’’ Pravda, Au-
ukraine 193
Democracy
gust 28, 2008; and ‘‘Russia to Sign Deal on Military Base in South Ossetia on Wednesday,’’ Ria
Novosti, April 6, 2010.
62. Oleh Protysk, ‘‘Majority-Minority Relations in the Ukraine,’’ p. 30. Protsyk went on in the
same quotation to say that this pledge has been kept by successive post-Orange cabinets.
63. From the fifteen countries that were once part of the Soviet Union, we therefore omit the
three Baltic countries and what is now called Moldova (then called Bessarabia), as they only became
part of the U.S.S.R. via the Molotov-Ribbentrop Pact of 1939. Unlike any of the eleven countries in
our set, all three of the Baltic countries had substantial experience with democracy in the interwar
period, and all have been members of the European Union since 2006.
64. For the raw data for 194 countries, see ‘‘The Freedom House Survey for 2009,’’ Journal of
Democracy 19, no. 2 (April 2010), pp. 142–143.
194 crafting state-nations
table 6.1
Percentage of ‘‘Confident Democrats’’ in Ukraine, Russia, and
Eight Central European Post-Communist Countries, 1996 and 2005
% of ‘‘confident democrats’’
1996 2005
Eight Central European post-Communist countries 59%
Russia 36
(Self-defined Russian nationality) (63)
Ukraine 30 63%
(Self-defined Ukrainian nationality) (63)
Sources: For the ten countries (eight—Bulgaria, Czech Republic, Slovakia, Hungary, Poland,
Romania, Slovenia, Croatia—compared to Ukraine and Belarus), see Richard Rose and
Christian Haerpfer, ‘‘New Democracies Barometer IV: A Ten Nation Survey,’’ Studies in Public
Policy, no. 262 (1996), p. 86. For Russia, see Richard Rose, ‘‘Getting Things Done with Social
Capital: New Russian Barometer VII,’’ Studies in Public Policy, no. 303 (1998), p. 44. For
Ukraine in 2005, see Richard Rose, ‘‘Divisions within Ukraine: A Post-election Opinion
Survey,’’ Studies in Public Policy, no. 403 (2005), p. 32.
Note: The definition Richard Rose uses for ‘‘confident democrats’’ is the same as he used in
the ten-country survey—that is, a respondent ‘‘disapproves suspension of parliament’’ (Q.29) and
‘‘considers suspension unlikely’’ (Q.28). The definition used for nationality is not essentialist but
self-definition. The question was, ‘‘What do you consider your nationality now?’’
68. ‘‘Change on the Horizon? Public Opinion in Ukraine Before the 2010 Presidential Elec-
tion,’’ International Foundation for Electoral Systems, November 2009.
69. For a more extensive discussion of the new concept, and the new practice, of ‘‘parliamen-
tarized semi-presidentialism,’’ see Alfred Stepan, ‘‘Introduction: Undertheorized Political Problems
in the Founding Democratization Literature’’ in Democracies in Danger, ed. Alfred Stepan (Bal-
timore: Johns Hopkins University Press, 2009), esp. pp. 9–14.
198 crafting state-nations
she controls a majority in parliament.70 However, classical theory and the modern
French experience are virtually silent on how the system should work if (as in
Yeltsin’s Russia) neither the president nor the prime minister has a majority.71 In
the case of Russia under Yeltsin—and in many of the post-Soviet cases—the presi-
dent, in the absence of a party majority, often rules on the margin of constitu-
tionality by decree or moves out of the democratic box entirely by closing the
parliament, tightly controlling the rewriting of a constitution in which presiden-
tial powers are greatly enhanced, and then holding a plebiscitary ratification of a
model that is most aptly called ‘‘super-presidential semi-presidentialism.’’72
Something like the above, except for the very important facts that no presi-
dent ever dissolved the parliament or was able to rewrite the constitution com-
pletely, occurred in post-independence Ukraine.73 There are two key points about
‘‘super-presidential semi-presidentialism’’ that are central to our inquiry as to
whether Ukraine can become a consolidated democratic state-nation. First, no
stable democracy in the world, certainly no European Union member country,
has a constitution that comes remotely close to being ‘‘super-presidential semi-
presidentialism.’’ Second, state-nations are facilitated if there is some degree of a
70. For an important new work on the theory of semi-presidentialism, see Cindy Skach, Borrow-
ing Constitutional Designs: Constitutional Law in Weimar Germany and the French Fifth Republic
(Princeton: Princeton University Press, 2006).
71. On these silences and the very exceptional conditions that allowed this model to function
reasonably well in France, see Alfred Stepan and Ezra Suleiman, ‘‘The French Fifth Republic: A
Model for Import? Reflections on Poland and Brazil,’’ in Stepan, Arguing Comparative Politics,
pp. 257–275.
72. See Timothy J. Colton and Cindy Skach, ‘‘The Predicament of Semi-Presidentialism,’’ in
Democracies in Danger. A shorter version of their argument appeared as ‘‘The Russian Predica-
ment,’’ Journal of Democracy 16 (July 2005), pp. 113–126.
73. For Kuchma’s augmentation of his powers, see Wolczuk, Moulding of Ukraine, pp. 205–
209. Oleh Protsyk, who wrote his 2000 Ph.D. dissertation at Rutgers on semi-presidentialism in
Ukraine, argues that both Russia and Ukraine are what Shugart and Carey would call ‘‘president-
parliamentary’’ and that Russia’s version gave the president more powers than in Ukraine. Nonethe-
less, he wrote that for most of his time as president, Kuchma, tried to ‘‘increase the formal powers of
the presidency’’ and that he ‘‘consistently played the role of a challenger to the existing constitutional
status quo.’’ See Oleh Protsyk, ‘‘Troubled Semi-Presidentialism: Stability of the Constitutional
System and Cabinet in Ukraine,’’ Europe-Asian Studies 55, no. 7 (2003), p. 1087. See also Andrew
Wilson, ‘‘Ukraine: Two Presidents and their Powers,’’ in Post-Communist Presidents, ed. Ray Taras
(Cambridge: Cambridge University Press, 1997). In the same volume, see the essay by Juan Linz,
‘‘Introduction: Some Thoughts on Presidentialism in Post-Communist Europe,’’ pp. 1–14. On at-
tempts to measure presidential powers in postcommunist Europe, see Timothy Frye, ‘‘A Politics of
Institutional Choice: Post Communist Presidencies,’’ Comparative Political Studies 30, no. 5 (1997),
pp. 523–552. Whereas Frye scored presidential powers in Slovenia, Slovakia, Latvia, the Czech
Republic, and Estonia as all in the 4.5 to 5.5 range, he gave Kuchma’s Ukraine a score of 15 and
Yeltsin’s superpresidential Russia a score of 16. Due to politics, however, unlike Putin’s Russia, there
was often gridlock under the first two directly elected presidents in Ukraine.
ukraine 199
74. Steven D. Roper, in his combined score of presidential powers—in each of the ten regimes
in Europe he classifies as ‘‘premier-presidential’’—gives Slovenia and Lithuania the lowest scores,
that is, he considers that they have the weakest presidents. See Roper, ‘‘Are All Semipresidential
Regimes the Same? A Comparison of Premier-Presidential Regimes,’’ Comparative Politics 34 (April
2002), table 3, p. 260. Roper does not give a score for Ukraine because he classified its regime as
‘‘presidential-parliamentary.’’ For the reduction of presidential powers in Poland and Portugal, see
Colton and Skach, ‘‘Predicament of Semi-Presidentialism.’’ For a more extensive analysis of ‘‘parlia-
mentarized semi-presidentialism’’ and details on the political conditions in Lithuania and Croatia
that made its adoption much more constitutionally rooted and consensual than in Ukraine, see
Stepan’s introduction in Democracies in Danger.
200 crafting state-nations
each other. The rivalry even helps to keep elections honest, since each side can
expose any attempt to falsify the ballot count.75
If somehow Ukraine were to evolve more toward the Portuguese, Polish, Slove-
nian, or Croatian version of parliamentarized semi-presidentialism, the Ukrainian
presidency would potentially be a ‘‘sharable’’ or ‘‘coalition-friendly’’ institution
and could become less a political source of ethnic polarization. Such a coalition-
friendly institution might help overcome the dual-government policy paralysis
and provide incentives for some state-nation style policy coalitions supported by
parties with different multinational configurations.
Such parliamentarized semi-presidentialism seems unlikely for Ukraine in the
foreseeable future. Viktor Yanukovych, who won the presidency in a close elec-
tion in February 2010, disliked the constraints on him in the semi-presidential
2004 Constitution. In July 2010 Yanukovych tried to get the two-thirds parliamen-
tary approval needed to hold a referendum on abolishing semi-presidentialism.
The opposition, and even two parties in his ruling coalition (the Communist Party
and the Lytvyn Bloc), refused to endorse a referendum.76 However, after four
Supreme Court judges resigned under what some of them complained was un-
due pressure from the Yanukovych administration, the Supreme Court unan-
imously declared the 2004 Constitution unconstitutional and reestablished the
1996 Constitution, with its much stronger presidential powers. Now the president
can select the prime minister and name cabinet ministers.
More than thirty years ago, in The Breakdown of Democratic Regimes, Linz and
Stepan argued that, even in once consolidated democracies, acts by democratic
incumbents can erode or even break down democracy. Ukraine has avoided
fragmentation by following some creative state-nation, as oposed to hard nation
state, policies. But the presidency of Yanukovych has weakened democracy. Most
observers said the October 31, 2010, municipal and regional elections were less
free and fair than the earlier 2010 presidential election. Reporters Without Borders
dropped Ukraine 42 places in its most recent world rankings of press freedom, and
there are increasing analogies between Putin and Yankovych. Yet Ukraine still has
more areas of political autonomy than Putin’s Russia. Democratic breakdown in
Ukraine is not over-determined. But democracy, as always, will have to be pro-
duced by struggle.
75. Henry E. Hale, ‘‘Ukraine: The Uses of Divided Power,’’ Journal of Democracy 21 (July 2010),
p. 86.
76. ‘‘Ukraine’s Yanukovych Fails in Move to Strengthen Powers,’’ Reuters, 10 July 2010.
chapter seven
Federacy
A Formula for Democratically Managing
Multinational Societies in Unitary States
The major subject of this book has been how states can democratically manage
multinational societies. We developed an ideal-type model, which we called a
state-nation, as a possible way to respond to this challenge, and then demonstrated
how some countries such as India and Spain have approximated empirically this
ideal type with relatively good results. A major component of our state-nation
model involved asymmetrical federalism.
However, we want to raise here a hard problem we have not yet discussed. Can
a unitary polity respond effectively to the nationalism of a territorially concen-
trated cultural minority within the state with its own ‘‘politically robust multina-
tional leaders’’ while nonetheless remaining a unitary and democratic state?
We will address this theoretical and political challenge in four steps.
First, we will identify the special nature of this problem and demonstrate why
and how unitary states do not normally have adequate political formulas available
to address this challenge.
Second, we will propose a political formula for how unitary states could demo-
cratically address this problem; we will call our newly conceptualized ideal-type
arrangements a ‘‘federacy.’’ We will advance a new ideal-type definition of fed-
eracy and its seven ideal-type institutional arrangements.
Third, we will shift from ideal-type analysis to comparative empirical analysis.
Have any actual unitary states resolved their problems with territorially concen-
trated cultural minorities by using policies that actually conform relatively closely
to our ideal type of federacy? Empirically, we will document why, and how, small
202 crafting state-nations
In keeping with our goal of expanding our imaginations beyond nation-states and
state-nations, let us first briefly hypothesize the character of the political problem
of territorially concentrated minorities in a unitary state.
Let us imagine an existing small or medium-sized independent state with a
unitary constitution, a relatively culturally homogeneous population, and a proud
nation-state sense of identity. Let us further assume that this otherwise well-
functioning unitary state has a territorially concentrated population that does not
identify with the history and culture of the unitary state. Its different identity will
most likely stem from one or more of the following six factors:
Let us make a further, not unreasonable assumption that this territorially con-
centrated minority population is unhappy with the status quo. At the very least,
they want more autonomy and self-governing arrangements that recognize their
culture and facilitate its development; some of the population may identify with
leaders and organizations who demand independence or seccession. This com-
bination of factors gives a multinational dimension to the unitary state.
Let us make some further constraining assumptions about two possible solu-
tions to this political problem: federalism or independence. Concerning federal-
ism, if the population and dominant political actors of the unitary state agree to
become an asymmetrical federal system, many of the problems of territorially
concentrated minority population, short of independence, could be addressed
reasonably well. However, if we assume that the vast majority of the state’s popula-
tion and its major political actors do not want to depart from what they perceive as
a well-functioning unitary state, we should assume that a federal state, much less
an option for an asymmetrical federal state, is highly unlikely.
Let us also assume that a full exit from the unitary state of the territorially
204 crafting state-nations
The minimal agreed set of arrangements for such a federacy, as an ideal type,
would have to satisfy the five institutional requirements.
Defining Characteristic 1: Federal-like division of state and federacy functions.
In order that the federacy, unlike other parts of the unitary state, can pass special
self-governing laws and administer their polity in such a way that it can address
many of the areas of greatest tension with the unitary state, there must be a classic
1. The reference, of course, builds on the reflections of Albert Hirschman’s Exit, Voice, and
Loyalty: Responses to Declines in Firms, Organizations, and States (Cambridge: Harvard University
Press, 1972).
federacy 205
bodies of the territorially concentrated population and the unitary state would
have to debate about, and vote for, the federacy arrangements. Legally, politically,
and often internationally, federacy arrangements would be much more binding
on the central government of the unitary state than devolution or decentraliza-
tion. The latter two might be unilaterally reversed by parliamentary majorities,
whereas a constitutionally embedded federacy would only be changed by mutual
supermajorities.2
Defining Characteristic 3: Existence of dispute resolution procedures. The fed-
eracy and the state, as part of the eventual autonomy agreement, would have
dispute resolution procedures about their respective powers and prerogatives.
Only in exceptional circumstances would a dispute go to the state’s highest court,
and this court would be able to make binding decisions for the entire polity only as
long as it does not in any way violate the constitutionally embedded autonomy
agreement. If the legislature of the center were considering a bill within its powers
that might have a special impact on the federacy, the federacy would often have
the right to present its views; on many matters, not only would consultation with
the federacy be required, but its consent as well.
Defining Characteristic 4: Reciprocal representation between the unitary state
and the federacy. The goal of the federacy arrangement is to create a high level of
trust, voice, and loyalty between the federacy and the center, so there is joint
citizenship in the federacy and the center. The citizens of the autonomous unit
would therefore be full citizens of the state, vote in state wide elections, and
have representatives in the parliament of the state as well as in their own par-
liament. The center would have an official representative in the federacy who
would also help coordinate those activities in the federacy that fall under central
state powers.
Defining Characteristic 5: The federacy is part of an internationally recognized
independent state. The autonomy agreement would be an internal law of the
unitary state and not part of international law. The autonomy agreement might be
derived from international agreements, but it is not part of international law
2. Some important constitutional theorists such as Geoffrey Marshall argue that political tradi-
tions and a sense of what is politically appropriate can give regular law de facto constitutionally
constraining qualities, even without written constitutional status. See Marshall, Constitutional
Conventions: The Rules and Forms of Political Accountability (Oxford: Oxford University Press,
1984). We will see such constitutional accretion in some of the cases we analyze, especially the
Denmark parliament’s acceptance of the de facto constitutional convention that its federacies had a
right to negotiate an exit from the state of Denmark. However, such conventions are normally only
produced over time, so our focus will be on written constitutionally embedded agreements.
federacy 207
(unless the state makes a specific exception) and would not create a ‘‘subject’’ of
international law.
If these five requirements are functioning, federacy arrangements may addi-
tionally contain and legitimate two peace-facilitating features that are not nor-
mally found in unitary or federal states.
Facilitating Characteristic 1: Role of international guarantors in the founding of
the federacy. A federacy arrangement, especially if it emerges in the context of
geopolitical conflict, may well combine elements of two of the greatest principles
in conflict since the late nineteenth century, the right of state sovereignty and the
right of self-determination of populations. In such cases, federacy arrangements,
more than in unitary or federal states, may involve some participation of con-
cerned neighboring states and international organizations in their emergence and
even in the establishment of some arrangements, particularly the distinctive cul-
tural rights of the federacy’s population.
Facilitating Characteristic 2: Role of the federacy in international treaties signed
by the center. More than in a federation, and totally unlike a unitary state, it is
possible that the very existence of a federacy may facilitate the federacy’s advocacy
and creation of some ‘‘opt-out’’ arrangements of treaties into which the unitary
state enters. This will be most likely if the leaders of the federacy, with the support
of the unitary state, believe that certain provisions of the treaty would be hurtful to
the preservation of the distinctive economy and way of life of the population that
the federacy had been created to protect. No other part of the unitary state would
have the status to construct such opt-out arrangements with the participation and
help of the center that normally negotiates them.
Given the defining characteristics of our ideal type of federacy, it should be clear
that it is quite different from the ideal type of a unitary state. This is so because of
the high degree of a federacy’s constitutionally embedded autonomy and preroga-
tives that are not enjoyed by other jurisdictions of the unitary state. However, as we
have seen, a unitary state can enter a federacy agreement; indeed, such an agree-
ment may be a particularly useful formula for managing a small or medium-sized
unitary nation-state’s problems with a territorially concentrated minority popula-
tion or nation.
208 crafting state-nations
3. Daniel J. Elazar, Federal Systems of the World: A Handbook of Federal, Confederal and
Autonomy Arrangements (Essex, UK: Longman Group, 1991), pp. xvi–xvii.
4. We say ‘‘much, but not all’’ because Bhutan has the right to reject Indian security advice. For
example, in 1961–62, during a time of great Sino-Indian tension, Bhutan unilaterally denied India
permission to guard the Bhutan-Tibetan border.
5. David A. Rezvani, ‘‘Federacy: The Dynamics of Semi-Sovereign Territories,’’ unpublished D.
Phil in Politics, University of Oxford, 2004.
210 crafting state-nations
fies cases of asymmetrical federalism as federacies. Thus, two of the four ‘‘fed-
eracies’’ he selects for special attention in an eleven-page appendix, along with the
Åland Islands and Greenland, are Catalonia and the Basque Country, neither of
which fits our definition of a federacy.6
We can insist on our five defining characteristics (and two facilitating conditions)
for a federacy in an ideal type; however, when we turn to the empirical analysis of
the contemporary world, is our ideal type an elegant but insignificant category in
world affairs, because there are few, if any, actual examples which approximate
the type? If there are examples, can we make cogent arguments about the condi-
tions in which they will tend, or tend not, to emerge? Can we make theory-
based predictions about the conditions in which federacies will function well or
function poorly? Here we begin with an examination of three cases; The Åland
Islands–Finland, the Faroe Islands–Denmark, and the Greenland-Denmark ar-
rangements. We will ask the following questions for each case; did a crisis con-
cerning exit or loyalty ever exist between a territorially concentrated minority
population and the relatively homogeneous unitary nation-state? Were policies
relatively close to some or all of our seven ideal-type federacy arrangements
created in response to the crisis? If they were created, what effect would we predict
they would have, and what effect did they have?
location also dominates the access to St. Petersburg, and is thus of great impor-
tance for the defense of three states: Sweden, Finland and Russia.’’8
The history of the islands reflects the contributing role of geography and
geopolitics as well as the tension between two powerful forces in international
politics—nationalism and the self-determination of peoples versus state building
and state maintenance. The institutional arrangements, their processes of enact-
ment, and the role of international actors, particularly the League of Nations,
exemplify how federacies can emerge.
From the Middle Ages until 1809, the Åland Islands were under Swedish rule
and were Swedish in language and culture. Finland was also under Swedish rule,
and the Åland Islands were administered by a governor from Abo, a city on the
Finnish coast. In 1809 Sweden was forced to yield Finland and the Åland Islands to
Russia. Within Imperial Russia, the Åland Islands were part of the Grand Duchy of
Finland, an entity that had a high degree of democratic self-government, with its
own parliament.9 Tore Modeen, a Scandinavian legal historian, describes how, in
the turmoil surrounding the Russian March Revolution of 1917, ‘‘just as the Finns
aimed at the independence of their country (including Åland), so the Ålanders in
turn wanted their islands to be detached from Finland and put under Swedish
sovereignty.’’10 In this standard work Modeen shows how, on the Åland Islands,
there had ‘‘arisen a strong separatist movement. . . . The Ålanders feared that
Finnish dominance when Finland gained its independence would be very strong.
They therefore looked upon union with Sweden as the surest guarantee for the
preservation of the national identity of the Åland Islands.’’11 Sweden encouraged
this prospect. Indeed, they sent a naval expedition in 1918, and Åland delegations
were well received in Sweden. Few authorities or contemporary observers who
have speculated on the subject doubt that if the principle of self-determination had
resulted in a plebiscite, Ålanders would have voted to join Sweden.12
8. Ruth Lapidoth, Autonomy: Flexible Solutions to Ethnic Conflicts (Washington, DC: U.S.
Institute of Peace Press, 1997), p. 70.
9. For example, in Finland in 1907 universal and equal suffrage was extended to all women and
men over 24 years of age. Thus Finland in 1910 had the highest percentage of voters in the 20-and-
over population of any of the fourteen West European countries documented in Peter Flora et al.,
State, Economy, and Society in Western Europe, 1815–1975: A Data Handbook in Two Volumes
(Frankfurt: Campus Verlag, 1983), pp. 93 and 108. The party in the 1910 elections with the highest
percentage of votes was the Social Democratic Party, p. 111. See also Osmo Jussila, From Grand
Duchy to Modern State: A Political History of Finland since 1809 (London: Hurst, 1999).
10. Tore Modeen, ‘‘Åland Islands,’’ in Rudolf Berhardt, ed., Encylopedia of Public International
Law, vol. 12 (Amsterdam: North Holland, 1990), p. 1.
11. Tore Modeen, ‘‘The International Protection of the National Identity of the Åland Islands,’’
Scandinavian Studies in Law (1973), pp. 178–179.
12. See in particular the sections ‘‘Swedish Occupation of the Aland Islands’’ and ‘‘Swedish
212 crafting state-nations
When we look at the six factors that we suggest might make a territorially
concentrated minority population feel quite distinct from the ethos of a unitary
nation-state, the Åland Islands in 1920 scored reasonably high on five of the six,
namely: (1) moderate but not great physical separation; (2) a different language;
(3) a moderate self-governing tradition they wanted to expand; (4) a belief that,
geopolitically speaking, their self-government could best be pursued by joining
Sweden; and (5) an economy and way of life sufficiently different that its residents
wanted to restrict ownership of property and most businesses to themselves. For-
tunately, there was no major history of warfare or coercive repression by Finland.
The absence of this feature, and the fact that both Åland and Finland were
parliamentary democracies, made the construction of a federacy feasible.
On the other side of the conflict however, the principle of state sovereignty and
state interests was also very strong. Finland objected to what it saw as an attack on
its sovereignty and territorial integrity, and insisted on maintaining the Åland
Islands as part of Finland. The French Prime Minister, Georges Clemenceau,
was particularly interested in strengthening the anti-Bolshevik cordon sanitaire
and thus wanted to back Finland which, unlike Sweden, had indicated its poten-
tial support for the cordon sanitaire.13 The U.S.S.R., for its own reasons, also
implicitly supported Finland because it wanted to maintain the demilitarization
of the Åland Islands and knew it had much more geopolitical-military leverage
over Finland than it had over Sweden, which was still a major regional power.
The international diplomacy involved in resolving the Åland controversy has
been well documented.14 However, in our judgment, a fundamental part of the
solution devised by the League of Nations involved federacy arrangements. Fol-
lowing the state sovereignty principle (and the pressure of the major victor powers
of World War I, mainly France, Britain, and the United States, all of whom
wanted to avoid more conflict and instability in the region), the Åland Islands
were awarded to Finland. But the price of the award of the Åland Islands to
Finland was limitations of a federacy nature on the Finnish unitary nation-state.
Numerous ‘‘self-determination of peoples’’ principles were built into the Åland
Agreement of 1921 and increasingly codified and consensually expanded in 1951
and 1991.15
Desires for a Plebiscite’’ in the extremely well-documented book by James Barros, The Aland Islands
Question: Its Settlement by the League of Nations (New Haven: Yale University Press, 1968), pp. 75–
84, 99–101.
13. Ibid., pp. 203–204.
14. Ibid.
15. See ‘‘Finland: Aaland Islands,’’ in Federal Systems of the World, ed. Elazar, pp. 98–100.
federacy 213
16. Modeen, ‘‘International Protection,’’ pp. 183–184. What is called the 1921 Agreement was
actually a decision that the Council of the League of Nations passed on June 24, 1921, with represen-
tatives of Finland and Sweden present. The government of Finland presented a bill to the Finnish
parliament containing the League of Nations provisions three days later. The first election of the
Åland parliament under their new prerogatives was not held until 1922. The 1951 Autonomy Act was
largely a response by the Finnish government and parliament to Åland Island demands since the
1930s for greater clarity about the various understandings and agreements that collectively were
called the 1921 agreement. After the end of World War II, the 1920s documents about Åland were
mutually negotiated between the federacy and Finland and were signed separately by the Finnish
parliament, the legislature of the Åland Islands, and finally by the president of Finland, see ibid.,
pp. 190–191.
214 crafting state-nations
of these state functions in Åland. Concerning civil defense, the evacuation of resi-
dents of Åland ‘‘may only be made with the consent of the Government of Åland’’
(Article 28). Nuclear energy is an exclusive function of the state of Finland—
‘‘however, the consent of the Government of Åland is required for the con-
struction, possession and operation of a nuclear power plant and the handling and
stockpiling of materials therefore in Åland’’ (Article 18). Telecommunications is a
state function—‘‘however, a state official may only grant permission to engage in
general telecommunications in Åland with the consent of the Government of
Åland’’ (Article 40).
Defining Characteristic 2: Quasi-constitutionally embedded political autonomy
of the federacy. No prerogative is worth much if the central state can unilaterally
reverse it. All constitutions have provisions for amendment, but the norm for
amendment is to require exceptional majorities. The Autonomy Act of Åland does
precisely this. The act stipulates that it can only be amended by a two-thirds
majority of both the Finnish parliament and the Åland parliament (Section 69).
Modeen is emphatic on the constitutional embeddedness of the Autonomy Act
and the restriction that neither Åland nor Finland can unilaterally alter the Auton-
omy Act: ‘‘The Autonomy Act cannot be repealed or amended without the ap-
proval of the Landsting [the Åland parliament]. Any such decision by the Fin-
nish parliament (Riksdag) shall follow the specific rules regulating constitutional
amendments.’’17 Lapidoth concurs, stressing that amending the Act requires the
joint action of Åland and Finland: ‘‘Collaboration or agreement between the
central authority and the province is indispensable.’’18
Defining Characteristic 3: Existence of dispute resolution procedures. This de-
fining characteristic of an ideal-type federacy is not met explicitly in the case of the
Åland Islands. The Act of Autonomy states that ‘‘if a conflict of authority arises
between Åland officials and State officials on a given administrative function, a
decision on the matter shall be rendered by the Supreme Court [of Finland]’’
(Section 60). However, the spirit and practice of the Act of Autonomy builds
in a type of dispute regulation. New state laws affecting only Åland cannot be
made, nor policies of the central state executed, in Åland—even in areas of state
responsibility such as civil defense and the development of nuclear energy and
telecommunications—without the express consent of the Åland legislature or gov-
ernment. As we shall see when we discuss treaties, although foreign policy is an
20. ‘‘Finland: Aaland Islands,’’ Federal Systems of the World, ed. Elazar, p. 87.
21. For a discussion of the exact opt-out provisions approved by the EC, see Lapidoth, Auton-
omy, pp. 237–238n38.
22. Modeen, ‘‘International Protection,’’ pp. 204–205.
218 crafting state-nations
asserted, ‘‘Because the Åland Islands have had autonomy for more than seventy
years and all those involved seem satisfied, it may be considered a successful case
of Autonomy.’’23
The Faroe Islands, if they had been an independent state in 2008, with its popula-
tion of about 50,000, would have constituted the eleventh least populous state in
the United Nations. The Islands, which have been linked to the Danish Crown
since the fourteenth century, are located between the Shetland Islands northwest
of Scotland and Iceland and are situated 1,300 kilometers from Copenhagen.24 As
Debes writes, ‘‘During the Middle Ages, Faroese developed into a separate Nordic
language. . . . No section of the Faroese people became Danish-speaking, while
Danish became the language of the authorities, in church, court, trade and the
administration at large.’’25
In the first constituent assembly in Denmark, three-quarters of the members
were elected by popular suffrage, but the Faroes were represented by a Crown
nominee. The new constitution, still without Faroese elected representation, was
applied to the Faroes in 1850, and the island was considered a province in the
unitary state of Denmark. The Faroes were given one seat in the Danish lower
house and one in the upper house and the right to an elected provincial assembly,
the Løgting, in its capital, Torshavn.
Competitive party divisions only really began early in the twentieth century.
‘‘The issue on which the first Faroese political parties were formed was whether to
maintain or loosen the constitutional link with metropolitan Denmark,’’ writes
West.26 The Unionist Party favored the former and the Home Rule Party the latter.
Due to the Unionist Party’s successful electoral appeals to the islanders’ worries
about losing valuable financial and cultural links to Denmark, the Unionist Party
was the dominant force in Faroese politics from 1906–36. The Social Democratic
Party was in the middle. It favored more home rule and greater language auton-
omy, but it never sought a complete break with Denmark because its social
policies were helped by the high welfare subsidies Denmark gave the Faroes.
However, electoral sentiments in favor of the equality of the Faroese language
with Danish and stronger home rule were growing by the late 1930s. At this point,
a context-changing geopolitical event happened. Germany occupied Denmark in
April 1940. The Danish king and government almost immediately capitulated to
Germany. All communications between Denmark and the Faroes were severed.
But the Faroes did not surrender. The day after the Danish surrender, the leader of
a new, more militant home-rule People’s Party in the Faroes, which had just won
six of the twenty-four seats in the Faroese legislature, ‘‘declared in the Lögting that
Faroe had passed from Danish sovereignty, and that the Lögting should now be
fully responsible for legislative and executive acts in the islands. This declaration
was coupled with the threat that if secession from Denmark had not been accom-
plished by 6 o’clock that evening, its proposers would themselves take whatever
steps were necessary.’’27
The People’s Party did not carry the day in the Løgting, but the legislature
passed a new provisional constitution that augmented Faroe’s capacity to govern
itself in virtually all domestic matters, while the friendly power, Britain, whose
protective presence was accepted, occupied the island and attended to the war.
After the 1943 elections, the nationalistic People’s Party was only one seat short of
an absolute majority. There was an economic boom owing to the Faroes’ key role
in providing fish to the hard-pressed British food market. In this context of isola-
tion from Denmark and growing self confidence in its self-governing capacities,
‘‘there was a remarkable intensification of nationalist agitation and the growing
support for the self-rule movement.’’28
At the end of the war, both the Danish and the Faroese authorities knew that
they could not go back to the prewar political arrangements, so negotiations
began. While the discussions were still going on, with the Danish government’s
proposal containing only modest new autonomy provisions, the People’s Party
argued for a plebiscite, as a way of informing their negotiators about the opinion of
the islanders. The moment of constitutional crisis in Faroe islands occurred
on September 1946. An inconclusive, Faroe-drafted plebiscite on independence
posed three alternatives: (1) ‘‘Do you want the Faroe Islands to secede from Den-
mark?’’ (2) ‘‘Do you want the Danish government’s proposal to be put into effect?’’
or (3) spoil your ballot if you want neither of the above but prefer a dominion
status short of complete independence. The second choice was favored not only
by the Unionists but also by the large Social Democratic Party, which wanted
autonomy but wanted maintenance of the traditionally large Danish welfare sub-
sidies to Faroe even more. The results in the September 14, 1946, vote were: 48.7%
for secession, 47.2% for the Danish government’s position, and 4.1% spoiled bal-
lots.29 Danish Prime Minister Knud Kristensen, who did not have a parliamentary
majority, took the view that the vote, though narrow for secession, should be
respected, and the Faroes were granted independence. However, the Danish
parliament, taking the view that the vote for independence was inconclusive, did
not support Kristensen. In the Faroes, the People’s Party now had a majority of
one and was able to pass a resolution to proceed with setting up a government
to negotiate independence.30 In Denmark ‘‘this resolution was widely seen as a
coup d’etat . . . the King, on the advice of the Danish government dissolved the
Løgting. . . . The Danish corvette, Thetis, was sent to Tórshaven . . . as a deterrent
to any popular disorder.’’31
For the Danish government to be seen as threatening force in the North
Atlantic against its own citizens was a national and international embarrassment.
For many in the Faroes also, a complete and costly break with Denmark was to
be avoided. New elections in the Faroes produced a coalition government: ‘‘In
March 1947 the Løgting voted unanimously in favor of a request to the Danish
government to grant legislative power to the Løgting, but on a basis of Faroe
continuing to be part of the Danish Kingdom. The result was the negotiation, over
the next twelve months, of the Faroese Home Rule Act. This was approved by
both the Danish Parliament and the Faroese Løgting.’’32
The Faroes, with their great physical separation from Denmark, separate lan-
guage, World War II legacy of de facto self-government, and fragile reliance on
low-technology fishing within their own shores and extremely vulnerable to out-
side fishing fleets, had four of the six features we have argued would make a
territorially concentrated minority population difficult to manage peacefully and
democratically within a unitary state. The negotiation of home rule was helped by
the complete absence of an irredentist power or any historical memory of actual
warfare or the use of armed coercive repression. But how close to our ideal type of
federacy was this Faroese Home Rule Act?
33. For a discussion of the division of powers see Lapidoth, Autonomy, pp. 113–114.
222 crafting state-nations
ment. Not being a state, the Faroes cannot join international intergovernmental
organizations.’’38
in almost twenty years an election was fought virtually without debate about
independence.40
If oil were ever discovered in the Faroes, and the islanders voted for indepen-
dence, the outcome would be completely unlike the 1946 crisis. There would be
no Danish corvette, no unilateral dissolution of the Faroe government by Den-
mark. Following the spirit of the Home Rule Act, a mutual ‘‘velvet divorce’’ would
be negotiated by two consenting democracies.
Greenland
40. See ‘‘Faroe Islanders Head to the Polls Saturday,’’ Agence France-Presse, January 17, 2008.
41. See Lapidoth, Autonomy, p. 144; and Frederick Harhoff, ‘‘Greenland’s Withdrawal from the
European Communities,’’ Common Market Law Review, no. 20 (1983), p. 18.
42. Harhoff writes that ‘‘Greenland has belonged to the Danish Crown since the fourteenth
century, but was first properly colonized by Denmark in the beginning of the eighteenth century
and administered by the Royal Greenland Trade Department in Copenhagen,’’ Ibid., p. 14.
43. Ibid.
226 crafting state-nations
by 52,000 persons, 42,000 of which are Eskimos of the same origin and culture as
the Eskimos of arctic Canada and Alaska.’’44
Economic differences from the unitary state and the desire for special protective
legislation. Greenland’s culture and way of living is exceptionally vulnerable:
‘‘Greenland is totally dependent upon its fishery. . . . Greenland’s fisherman fish
exclusively within Greenland’s own fishing territory, thus having no interest in
fishing in foreign waters.’’45 Because Greenlanders are so dependent on fishing off
their own shores, they feel very threatened by the prospect of allowing other
countries to fish in their waters. A common market agreement by the unitary state
that would allow other states to fish off Greenland’s shores, while also permitting
Greenlanders to fish thousands of miles away off member states’ shores, is thus
quite threatening.
Three things happened to put pressure on Denmark’s colonial arrangements
concerning Greenland. First, like the Faroes, Greenland was completely cut off
from Denmark by the German occupation in World War II. Being so close, the
United States, and not Britain, became the occupying country during the war.
Second, in 1946, in the postwar atmosphere of de-colonization, Denmark in 1946
felt obliged to acknowledge to the United Nations that Greenland was a ‘‘non-self-
governing territory.’’46 Third, in November 1947 a new Social Democratic govern-
ment came to power in Denmark. The Social Democrats were not comfortable
with Denmark’s classification by the UN as a colonial power subject to routine
reports and inspections, so in 1953 the government decided to alter Denmark’s
colonial power status by making Greenland a full voting member of the unitary
state of Denmark.47
Before colonization, Greenland had a robust tradition, which Elazar calls
‘‘tribal republicanism.’’48 However, at the time of incorporation into Denmark as a
normal voting province in 1953, Greenland had a very low level of political
mobilization and no political party life. The trigger to mobilization and the devel-
opment of a party system involved the European Community. As Axel Sørensen, a
historian of Greenland nationalism, explains, ‘‘The Danish entry into the Com-
mon Market in 1972 was the clearest single impetus which got the political de-
velopment in Greenland moving. A referendum was taken all over the realm.
44. Ibid.
45. Ibid., p. 22.
46. Ibid., p. 14.
47. See Axel Sørensen, ‘‘Greenland: From Colony to Home Rule,’’ in Ethnicity and Nation
Building in the Nordic World, ed. Sven Tägil (London: Hurst, 1995), pp. 98–102.
48. ‘‘Finland: Aaland Islands,’’ Federal Systems of the World, ed. Elazar, p. 80.
federacy 227
autonomy from Denmark. The referendum passed with approximately 75% of the
vote.54 Negotiations ensued between Greenland and Denmark about what were
the autonomies for which Greenland would like to assume full financial and po-
litical responsibility. The new ‘‘Act on Greenland’s Self Government,’’ an agree-
ment between the Greenland government and the Danish government as ‘‘equal
partners,’’ dated June 2009 and signed by Denmark’s Queen Margrethe II, ex-
plicitly recognized in the first sentence of the preamble that ‘‘Greenland is a
people pursuant to international law with the right of self-determination.’’55 Arti-
cle 8 of the act goes on to say, ‘‘Decision regarding Greenland’s independence
shall be taken by the people of Greenland.’’ Article 3 of the act stipulates that,
until such time as there is independence, Denmark will continue to give Green-
land an annual subsidy of 3,439.6 million kronor per year (approximately 633
million USD, or approximately $11,000 per capita). This sum will decline as
revenues from Greenland’s mineral resources increase. Greenland’s competen-
cies in arranging international agreements with the consent of Denmark were also
increased in the 2009 agreement.
beyond scandinavia?
Can the concept and some of the practices of federacy travel beyond the small,
relatively peaceful democracies we have explored so far? What is the scope value
of our concept? Can it travel into much more conflictual situations? Can it travel
to non-Western, especially Islamic, cultures? In the second half of this chapter, we
examine four different cases. Might federacy be useful in some transitions to
democracy after authoritarianism or defeat in war? For this we examine Italy in
the immediate aftermath of World War II, when the unitary state had a ‘‘politically
robust multinational’’ situation in which five peripheral regions either had leaders
and movements who wanted to exit or were demanding greater autonomy. In less
than a decade, Italy was a consolidated democracy with no secessionist move-
ments. Why? How? Our second case is Portugal in 1974, when the Carnation
Revolution broke out in one of Europe’s oldest unitary states. Why and how did
the constituent assembly defuse a secessionist movement on the Azores? Third is
the case of Corsica. The basis for our ideal type of a normatively grounded,
unitary nation-state model is, of course, France. Why did the National Assembly
54. Alan Cowell, ‘‘Greenland Vote Favors Independence,’’ New York Times, November 26,
2008, available at www.nytimes.com/2008/11/27/world/europe/27greenland.html, accessed No-
vember 30, 2008.
55. Denmark. Act no. 473 of June 12, 2009, Act on Greenland Self-Government.
federacy 229
try to alter this model in the case of Corsica in 2001? Fourth, and last, concerns
Indonesia. Can a federacy be constructed between two opponents in an ongoing
civil war? Can federacy travel well even to a Muslim context, which is often
alleged to be particularly hostile to shared sovereignty? We will look at Indonesia
and Aceh to explore these last two questions. Let us begin our exploration of the
scope value with Italy after World War II.
Italy’s economy, politics, and unitary state at the end of World War II were all
facing crises. Due to Mussolini’s alliance with Hitler, Italy’s international and
domestic legitimacy was dangerously low. In the country’s peripheries, territo-
rially concentrated minority populations, in five different areas, partly in response
to Mussolini’s authoritarian centralizing policies, had leaders and movements
seeking exit or self-government from the Italian state. These areas were: German-
speaking South Tyrol, where the vast majority of the population wanted to rejoin
their irredentist neighbor, Austria; the French-speaking province of Valle D’Aosta
on the alpine border with France, which had recently sent troops and agents to the
province, making annexation a live issue; Friuli-Venezia Giulia on the border
with Yugoslavia, with many Slovene speakers, much of the territory of which the
Allies’ war partner Tito was occupying and claiming as part of Yugoslavia; and the
islands of Sicily and Sardinia, with their very distinctive social structures and
problems.56 In terms of the categories we have developed in this book, Italy in
1944–45 was a ‘‘politically robust multinational polity’’ governed by a unitary state.
South Tyrol, in what is now the province of Bolzano, is adjacent to the strate-
gically valuable Brenner Pass connecting the Alps with northeastern Italy. Before
World War I this territory was part of Austria, with a population that was approxi-
mately 86% German-speaking, 8% Italian-speaking, and 4% Ladino-speaking.57 In
1919, in the Treaty of St. Germain, the territory was given to Italy as part of a secret
agreement (later published) for joining the allies in World War I. Mussolini’s
fascist regime imposed a particularly harsh and authoritarian set of unitary-state
and Italian nation-building policies on South Tyrol. The German schools were
closed, Italian was made the only official language, there was extensive state-
56. For an invaluable map showing these five peripheral regions and the fifteen other regions of
Italy, see ‘‘Italy,’’ Federal Systems of the World, ed. Elazar, p. 121.
57. For rapidly changing language estimates for the province of Bolzano in 1910, 1921, 1939, 1943,
1946, 1951, 1953, and 1961, see Anthony Evelyn Alcock, The History of the South Tyrol Question
(Geneva: Graduate Institute of International Studies, 1970), p. 497, table D.
230 crafting state-nations
sponsored Italian immigration into the area, and the name of the territory was
changed to Alto Adige. In 1926 Mussolini proudly announced: ‘‘The Germans of
the Alto Adige are not a national minority but an ethnic remnant.’’58
Not so. German speakers were still the largest and most politically organized
part of the population in South Tyrol (Alto Adige) in 1945. Almost immediately
after the war, a German-speaking party, the South Tyrolean People’s Party (Süd-
tiroler Volkspartei, SVP), was formed and rapidly became the dominant social
movement and political party in South Tyrol. In the name of self-determination,
the SVP demanded a plebiscite over their status. With the strong domestic politi-
cal and international diplomatic irredentist support of Austria, the SVP’s goal was
to exit Italy and accede to Austria.59 The Italian-speaking minority in South Tyrol
and the Italian-speaking majority in Italy were in turn deeply suspicious of the
German-speakers of South Tyrol. Hitler and Mussolini had arranged for a plebi-
scite of sorts in South Tyrol in which German-speakers could opt to retain Italian
nationality or migrate to Hitler’s Reich and receive German citizenship. Out of a
total number of the 229,500 inhabitants of Bolzano entitled to opt out of Italy and
to migrate to the Reich, 166,488 did so.60
A second troubling area for the unitary state was the Alpine province of Valle
d’Aosta, situated in northwestern Italy, bordering France, Switzerland, and Pied-
mont, the powerhouse of Italian unification in the nineteenth century.61 Histori-
cally, Valle d’Aosta had had a French-speaking population, which in return for
pledging loyalty to the Savoias in 1191, received special ‘‘rights and privileges and
some self government’’ and gradually developed substantial identity with Italy.62
Up until unification in 1860, if the question ‘‘Can you be Italian while continuing
58. Rolf Steininger, ‘‘Back to Austria? The Problem of South Tyrol in 1945/46,’’ European
Studies Journal 7, no. 2 (Fall 1990), p. 52.
59. For an excellent analysis of the SVP from 1945 to the present, see Günther Pallaver, ‘‘The
Südtiroler Volkspartei: From Irrendentism to Autonomy,’’ in Autonomist Parties in Europe: Identity
Politics and the Revival of Territorial Cleavage, vol. 2, ed. Lieven De Winter, Margarita Gómez-
Reino, and Peter Lynch (Barcelona: Institut de Ciències Polítiques i Socials, 2006), pp. 161–188.
60. For an thorough treatment of this episode, see Alcock, ‘‘Hitler, Mussolini, and the Options
Agreement of 1939,’’ in History of the South Tyrol Question, pp. 45–59. On the immensely compli-
cated and fraught national and international negotiations on the rights of the 70,000 ‘‘optants’’ who
migrated to the Reich to return to Italy as citizens, and thus as voters in the South Tyrol, see
‘‘Autonomy and Options: October 1946–February 1948’’ in the same volume, pp. 148–194. See also
Steininger, ‘‘Back to Austria?’’ p. 54.
61. For an important account of the Italian unification process, and particularly the central role
of Piedmont, see Daniel Ziblatt, Structuring the State: The Formation of Italy and Germany and the
Puzzle of Federalism (Princeton: Princeton University Press, 2006).
62. See Marco Cuaz, ‘‘La Valle d’Aosta. Un’identità di frontiera fra Italia, Europa ed etno-
nazionalismi,’’ in Altre Italie: Identità nazionale e Regioni a statuto speciale, ed. Gaspare Nevola
(Rome: Carocci editore, 2003), pp. 1–18, esp. pp. 1–3.
federacy 231
to speak French?’’ was asked, the most likely voluntary answer in Valle d’Aosta
probably would have been, ‘‘I am an Italian who speaks French.’’63 Such an
answer would have been a clear illustration of what we have called multiple but
complementary identities. Unification of Italy, with its increased taxes and greater
centralization, eroded these pro-Italian sentiments somewhat, but it was Mus-
solini’s Tyrolean-type, authoritarian Italianization policies that turned these mul-
tiple identities into polar identities. Within Valle d’Aosta, sentiments in favor
of joining France emerged and were encouraged by irredentists in France. As
Kogan’s exemplary political history of early postwar Italy states: ‘‘In April 1945,
French troops crossed the Alps to occupy adjacent territory in Piedmont and the
Val d’Aosta. French agents tried to win the French-speaking Valdostani over to the
idea of annexation.’’64 At a meeting of the Council of Foreign Ministers of the
victorious allies, France gave up some, but not all, of its demands. A lot was at
stake for potentially fragmenting Italy: ‘‘The French demands, if realized, would
bring France down onto the Italian side of the Mountains and eliminate the
natural barriers to invasion routes into the Po Valley. Also on French soil would be
some major hydroelectric power installations that provided substantial portions of
electric current to such cities as Genoa and Turin.’’65
In the atmosphere of a weakened Italian state, facing since 1943 a resurgent
communist movement aiming at control of the Italian polity, separatist move-
ments emerged in Sicily and Sardinia. Sicily had the third-largest population
of any Italian province and a history of pre-unification autonomy and post-
unification conflict with the unitary state of Italy. Sardinia was once the proud
Kingdom of Sardinia. Kogan writes: ‘‘Separatist organizations had begun to oper-
ate in . . . Sicily after the liberation of these two major islands of Italy. Conservative
and reactionary forces feared that radicals would take over the central government
on the mainland. The more substantial landowners believed they could better
protect their holdings by separation. Although in 1944 the Americans and British
repudiated the separatists’ claims of Allied support, the agitation nevertheless
continued. Its peak was reached in the immediate postwar period.’’66
The fifth peripheral area in the postwar conflicts was on the border with
Yugoslavia. Yugoslavia was a victor country in World War II on the side of the
Allies. Tito used Mussolini’s repression of minorities to fan Slovene and Slav
desires to join Yugoslavia. Tito gave the fragile Italian government a major blow
by occupying many areas with Italian populations. Italy was particularly worried
about losing much of the Friuli-Venezia Giulia area, especially the great port city
of Trieste, which had an important Slovene minority.67
How was Italy to respond to these five simultaneous challenges to its unitary-
state authority? The political history of unified Italy did not offer much in the way
of a useable past. In theory, federalism, especially asymmetrical federalism, would
have seemed to have offered the most possibilities. However, Italy, which, despite
its pre-unification character, was divided into five various kingdoms and a papal
state, a situation that might naturally have led to German-style confederation
morphing eventually into a federation, in fact became a unitary state in 1860.
None of Italy’s major unification leaders, Cavour, Garibaldi, or Mazzini, were
sympathetic to federalist ideas. In any case, the largest political force in 1945 in
Italy were the Christian Democrats. Their major challenger was the Communist
Party of Italy. While in theory, the Christian Democrats were in favor of decentral-
ization, in reality they feared passing power to the regions. The regions of the
famous ‘‘Red Belt’’ of central Italy—such as Umbria, Emilia-Romagna, Tuscany,
and to some extent Marche—were the strongholds of the Communist Party. In
fact, while the constitution of 1948 called for regionalization and decentralization,
formal bills to transfer more power to the regions were not passed until 1970.
Actual power transfers were only implemented in the late 1970s, thirty years after
the crisis in the five peripheral regions described above.68
So, if the government of the unitary state had neither the will nor the time to
decentralize, and if full federalization, much less asymmetrical federalism, was
not on the agenda, what could the Italian government do in a hurry? The answer,
in essence, was to create de facto federacies (called ‘‘regions of special statute’’)
with enhanced cultural and political prerogatives as rapidly as they could. In the
case of Valle d’Aosta, this was done by a decree law on September 7, 1945, to blunt
French territorial claims and, in Sicily, on May 15, 1946, to defuse the separatist
movement. These decrees were later made Constitutional Laws on February 20,
1948, the same day that Trentino Alto Adige and Sardinia were also made federa-
cies by Constitutional Laws.69 The Italian constitution of January 1, 1948, in
Article 116, divided Italy into twenty regions and created the category of ‘‘regions of
special statute’’ for five regions. The five peripheral regions were made ‘‘regions of
special statute’’ with separate powers and prerogatives spelled out in their special
statutes.70 The fifteen other regions were ‘‘normal’’ regions with identical pow-
ers and no special prerogatives. The goal of the special statutes—which we call
federacies—in the words of one of its intellectual authors, Federico Chabod, was
to establish territorial autonomy without encouraging an ethnic minority that was
‘‘critical and distinct from the Italian people.’’71 All five crises were substantially
defused, and a degree of multiple but complementary identities was gradually
restored. Italy, though still a unitary state, had completed its transition to democ-
racy by the late 1940s.72
The most complicated case was that of South Tyrol, and this case tells us a
lot about some of the special international and national arrangements that
can be negotiated once a context of a federacy is on the table. The origin
of the special statute for South Tyrol was not in a constituent assembly but
in the De Gasperi–Gruber Accord signed by the foreign ministers of Italy and
Austria on September 5, 1946, and linked to the Paris Peace Treaty (which in-
volved the United Nations).73 Some of the specific commitments were expanded
and constitutionalized in 1948 and 1972 in further special statutes. The De
Gasperi–Gruber Accord was helpful in defusing the crisis, especially in the
area of equal cultural, political, and employment rights for German-speaking
citizens. Some of the key agreements that allowed Karl Gruber, the Austrian for-
eign minister, to stress that the accord would significantly improve the rights
of the German-speaking former Austrian citizens in South Tyrol, were the
following:
70. The indispensible book on these special statute regions, and how each special statute, in
different ways, helped defuse the crises we have discussed, is Nevola, ed., Altre Italie.
71. Cuaz, ‘‘La Valle d’Aosta,’’ p. 7.
72. For a classic study of the Italian democratic transition, see Gianfranco Pasquino, ‘‘The
Demise of the First Fascist Regime and Italy’s Transition to Democracy, 1943–1948,’’ in Transi-
tions from Authoritarian Rule: Southern Europe, ed. Guillermo O’Donnell, Philippe C. Schmitter,
and Laurence Whitehead (Baltimore: Johns Hopkins University Press, 1986). For excellent analyses
of democratic consolidation in postwar Italy, see Maurizio Cotta, ‘‘Elite Unification and Demo-
cratic Consolidation in Italy: An Historical Overview’’ in Elites and Democratic Consolidation in
Latin America and Southern Europe, ed. John Higley and Richard Gunther (Cambridge: Cam-
bridge University Press, 1992), pp. 146–177; and Leornado Morlino, Democracy between Consoli-
dation and Crisis: Parties, Groups, and Citizens in Southern Europe (Oxford: Oxford University
Press, 1998).
73. The full Italian version of the accord, and a full English language translation, is found in
Andrea Di Michele, Francisco Palermo, and Günther Pallaver, eds., 1992: Fine di Un Conflitto:
Dieci anni dalla chiusura della questione sudtirolese (Bologna: il Milano, 2003), pp. 77–79. This
accord was an appendix to the Treaty of Paris.
234 crafting state-nations
The accord enabled Austrian political leaders not to be too adamant in their
irredentist drive. It also enabled Alcide De Gasperi, the Italian foreign minister, to
stress that the major concessions to German-speakers would strengthen Italy’s
frontiers and protect its electrical resources while ensuring the language rights of
the Italian minority in South Tyrol. These and many other issues were included
in The Special Statute for the Trentino–Alto Adige as Constitutional Law of
February 2, 1948.
This Constitutional Law meets our embeddedness requirement for a federacy.
Article 88 of this statute stresses its constitutional embeddedness by stating: ‘‘For
the amendment of the present law the procedure established by the Constitution
shall be applied.’’ Articles 4 through 17 spell out the federal-like constitutionally
embedded powers of the region. Citizens of Bolzano are full citizens of Italy, and
since 1948 have received six seats in the Italian parliament, four in the lower house
and two in the upper house. As in the Scandinavian federacies, the unitary state is
represented in the federacy by a commissioner.74
Notwithstanding these gains, there was still a major point of contention in the
new federacy. It was solved in a novel form of dispute settlement involving sub-
75. See Dietrich Schindler, ‘‘South Tyrol,’’ in Encylopedia of Public International Law, vol. 12,
ed. Rudolf Berhardt (New York: North-Holland, 1990), pp. 348–350; and R. Steininger, ‘‘75 Years
After: The South Tyrol Conflict Resolved; A Contribution to European Stability and a Model for
Settling Conflicts?’’ in Austria in the 1950s, ed. Gunter Bischof, Anton Pelinka, and Rolf Steininger
(New Brunswick, NJ: Transaction Publishers, 1995), pp. 116–137.
236 crafting state-nations
Italian parliament and to the European Parliament, and they staff much of the
public administration in South Tyrol. But, very importantly, the SVP has high
prestige with all three language groups. Günther Pallaver, in his review of late
1990s public opinion surveys, reports that 78% of the German-speaking voters
responded that they were happy with the work of the regional government, but sig-
nificantly, an impressively high 62% of Italian speakers and 60% of Ladin speakers,
also expressed satisfaction with the regional government, which of course was run
by the SVP. All three language groups also answered that they saw advantages in
autonomy (German 83%, Ladin 68%, and Italian 60%).76
Nationally and internationally, the SVP was an active, and fairly positive,
political presence. The SVP maintained good relations with most parties in the
Italian parliament, especially the Christian Democrats, who were the dominant
party in the first three decades of the Italian democratic transition. The SVP also
maintained close contacts with the Austrian governments of the day. In Austria,
public posts are distributed proportionally to party members of the two major
parties, the Christian Democrats and the Social Democrats, in a system known as
proporz, so they understood South Tyrol’s politics of ‘‘language proporz’’ for all
public posts.
The conflict in Valle D’Aosta was settled much more rapidly but along similar
lines. The Allied Powers wanted a strong postwar Italy but could not come out
against French demands until Italy publically reversed all of Mussolini’s anti-
French policies. Italy rapidly did this with the September 7, 1945, Decree Law
setting up the federacy we discussed earlier. The federacy’s new language poli-
cies, combined with the well-supported and self-administrating French-language
schools, a robust economy helped by subsidies built into the special statute, and
large tourist and hydro-electric investments, turned Valle D’Aosta into a multi-
ethnic frontier province with one of the highest standards of living in Italy.77
The peak of separatist demands in Sardinia and Sicily was right before the
granting of their special statutes.78 In the judgment of Daniel J. Elazar, ‘‘Sicily and
Sardinia had . . . active secessionist movements which the general government
successfully defused through the granting of special regional status.’’79 In Sicily,
one observer asserts that Sicilian regionalism ‘‘concentrated itself in the separatist
phase in the first half of the 1940s.’’80 In the case of Sardinia, this would also seem
to be the case, because military observers estimated that in 1944–45, the autono-
mist party was the largest party in Sardinia, with forty thousand members, but
experienced rapid decline in the 1950s.81
The peripheral crisis with Slovenians and Yugoslavia took longer to solve.
Italy did lose substantial Italian-speaking areas to Yugoslavia but was eventually
awarded Trieste after it agreed to protect Slovene language rights and put credible
guarantees to this effect in the 1963 special statute.82
Portugal
Portugal is one of the relatively few nation-states in the world with no national,
ethnic, linguistic, or ‘‘native’’ minorities. Its borders date back centuries, and it has
no irredenta (except for a claim to the border city of Olivenza it lost to Spain in the
eighteenth century). It has always been a unitary state. Even recently, the citizens
overwhelmingly rejected in a referendum a reform that would have represented a
possible move toward federalism precisely because they have historically con-
ceived their state as a unitary nation-state.
However, in the archipelago of the Azores, 800 miles west of Lisbon, during
the turmoil of the 1974 Portuguese Revolution, also called the Carnation Revolu-
tion, secessionist sentiments emerged. While the constitution-makers were work-
ing, many people in Portugal feared that the radical junior officers who had
started the revolution in April 1974 might end up having the power to install a
communist regime. The secessionist sentiments in Azores were driven by these
fears but also had a special contributing geopolitical dimension. The United
States had a major strategic air base in the Azores, Lajes Field, which was co-
administered by Azoreans and gave them substantial employment. This, and the
close ties to Azorean immigrants in the United States, strengthened anticommu-
nist sentiments, which in turn led to secessionist activities. Opello’s standard
history of the Carnation Revolution analyzes its separatist impact in the Azores:
‘‘Powerful separatist tendencies . . . appeared on the islands in 1974 and 1975 when
it seemed that Portugal would likely have a communist government. In those
years, the desire to separate was given voice by the . . . Azorean Liberation Front
80. See Christophe Roux, ‘‘The Partito Sardo D’Azione: Regionalist Mobilization in Southern
Italy,’’ in Autonomist Parties in Europe, ed. De Winter, Gómez-Reino, and Lynch, p. 191.
81. Ibid., p. 211.
82. See Kogan, Political History of Italy, pp. 3–6.
238 crafting state-nations
Article 6 of the constitution states that 1) ‘‘the state is a unitary one organized to
respect the principles of autonomy of local authorities and democratic decentral-
ization of the administration; and 2) the archipelagos of Azores and Madeira
constitute autonomous regions with their own political and administrative status
and self-governing organs.’’ This formula fits well within our concept of federacy.
What for our purposes is particularly important is that in Article 288, paragraph
O, the constitution lists which matters can only be changed by amendment and
therefore require exceptional majorities. The new federal arrangements were
given such constitutionally embedded status: ‘‘The laws revising the constitution
safeguard . . . the political and administrative autonomy of the archipelagos of the
Azores and Madeira.’’
Our reading of the constitutional norms leads us to consider Azores as a clear
83. Walter C. Opello Jr., Portugal: From Monarchy to Pluralist Republic (Boulder, CO: West-
view Press, 1991), p. 16, quotation from p. 153. For U.S. policies in this period, see Kenneth Maxwell,
‘‘The Thorns of the Portuguese Revolution,’’ Foreign Affairs 54 (January 1976), pp. 250–270.
84. The same provisions were extended to the archipelago of Madeira, but more for reasons of
political symmetry than political threat. Because the constitutional writers’ fundamental concern
was the Azores, we will restrict our analysis to the Azores.
federacy 239
case of a federacy. However, we also think that, given the unitary conception of
the state in Portugal, and their otherwise strong nation-state character, it is doubt-
ful, with no secessionist dangers or foreign threat, that a federacy in Portugal is as
necessary now as in the other cases we have considered.85
Corsica is an island in the western Mediterranean that had a brief period of virtual
independence (1755–69) but was eventually sold by Genoa to France. To this day
the vast majority of the total population of 260,000 inhabitants of Corsica speak
Corsican, a language which is closer to Italian than to French. In a 1982 survey of
those inhabitants who defined themselves as of Corsican origin, 96% said they
understood Corsican, and 86% said they regularly spoke it.86 The ‘‘administrative
and legal role [of Corsican] is minimal for it has no official status.’’87 This linguis-
tic marginalization of the majority was a longstanding source of tension on the
island. However, sustained violence with autonomist dimensions began in the
early 1970s. Between 1971 and 1998 there were 45 political murders in Corsica, 21
of which were explicitly claimed by nationalists.88 In 1995 and 1996 there were
1,172 bomb attacks. On February 6, 1998, Claude Erignac, the prefect of Corsica,
the senior French official in the governing structure of France on the island, was
assassinated.
The origins of the political violence have a temporal relationship to two major
policy initiatives, taken near-unilaterally by the French state, that directly al-
located scarce resources in Corsica to non-Corsicans. Daftary describes these
French state policies thus: ‘‘A major reorganization of the agricultural sector
introduced in 1957 . . . failed to bring any benefits to the local population for it was
used to relocate 17,000 pied-noir colonists from Algeria who were given preference
85. For an overall view of the turmoil of the Portuguese political process in 1974–1976, see the
classic work by Kenneth Maxwell, ‘‘Regime Overthrow and the Prospects for Democratic Transi-
tion,’’ in Transitions from Authoritarian Rule: Southern Europe, ed. O’Donnell, Schmitter, and
Whitehead, pp. 109–137. For the placement of the Portuguese democratic transition in comparative
perspective, see Juan J. Linz and Alfred Stepan, ‘‘From Interim Government to Simultaneous
Transition and Consolidation,’’ in Problems of Democratic Transition and Consolidation, ed. Linz
and Stepan (Baltimore and London: Johns Hopkins University Press, 1996), pp. 116–129.
86. Farimah Daftary, ‘‘Insular Autonomy: A Framework for Conflict Resolution? A Compara-
tive Study of Corsica and the Alland Islands,’’ Global Review of Ethnopolitics 1, no. 1 (September
2001), pp. 19–40, data on p. 24.
87. Ibid.
88. J. L. Briquet, ‘‘Le problème corse,’’ Regards sur l’actualité, no. 240 (April 1998), pp. 25–37.
240 crafting state-nations
in obtaining land. The state policy of development of tourism starting in the late
1960s was also perceived as colonial expansion.’’89 In the judgment of Marc Smyrl,
‘‘The rise of militant modern separatist movements since the 1960s in Corsica is
linked directly to’’ such policies. ‘‘The first violent manifestations of separatist
sentiment in Corsica came with bomb attacks against pied noir property.’’90
In 1997 a Socialist prime minister, Lionel Jospin, came to power in France and
indicated his government’s interest in opening a dialogue about ways to reduce
conflict in Corsica. Before Jospin’s arrival in Corsica in September 1999, a process
of internal political unification had begun in Corsica, and the center-right presi-
dent of the Corsican Assembly publically called for political autonomy.91 A seven-
month process of negotiation between the French government and representa-
tives of all parties to the Corsican Assembly began, including the deputies from
Corsica Nazione, the only separatist organization in the Assembly. The result, in
the summer of 2000, was the crafting of a concord, to be implemented gradually
over four years, that Smyrl argues was ‘‘accepted by all but the most intransigent
forces on each side.’’92 By May 2001, the French Assembly had passed a bill which
contained, strikingly for the French unitary nation-state, certain asymmetrical,
federacy-like features. In the judgment of Smyrl, ‘‘in a significant departure from
past French practice’’ the Assembly of Corsica was given the ‘‘power to pass
secondary legislation—that is to adapt national laws to the special conditions of
the island.’’ In addition, there were ‘‘provisions encouraging the teaching of the
Corsican language in public schools and generally reinforcing Corsica’s cultural
autonomy.’’93
The key point we would like to make is that the most important democratically
elected body of French representatives, the French National Assembly, after a
number of years of reflection and negotiation, decided that political practices and
principles restricted to a nation-state repertoire were not appropriate for the peace-
ful and successful management of French-Corsican relations. A federal alterna-
tive, especially of the full state-nation sort, was not on anyone’s political agenda. In
94. Smyrl, ‘‘France,’’ p. 218. For the exact wording of the council, see France; Constitutional
Council, Decision 2001-454 DC of January 17, 2002, paragraphs 18–21.
95. France; Constitutional Council, Decision 2001-454 DC of January 17, 2002, paragraph 29.
96. For two well-documented articles on the still fragmented nature of party politics in Corsica,
see Isidre Molas, ‘‘Partis nationalistes, autonomie et clans en Corse,’’ and Thierry Dominici, ‘‘Le
système partisan nationalitaire corse contemporain: étude d’un phénomène politique,’’ both in the
Barcelona journal, BCN Political Science Debates 4 (2005), pp. 5–42 and 43–88, respectively.
242 crafting state-nations
97. The best documentation and analysis of Aceh’s conflicts with Indonesia is found in Edward
Aspinall, Islam and Nation: Separatist Rebellion in Aceh Indonesia (Stanford: Stanford University
Press, 2009). Also see his ‘‘Modernity, History, and Ethnicity: Indonesian and Acehnese Nationalism
in Conflict,’’ Review of Indonesian and Malaysian Affairs 36, no. 1 (2002), pp. 3–33. For a convincing
argument that for GAM, nationalism eventually became more important than Islamism, see As-
pinall, ‘‘From Islamism to Nationalism in Aceh, Indonesia,’’ Nations and Nationalism 13, no. 2
(2007), pp. 245–263. Also useful on GAM is Anthony Reid, ed., Veranda of Violence: The Background
to the Aceh Problem (Singapore: Singapore University Press, 2006). For a detailed analysis of the
military and ideological dimensions of GAM as a separatist movement, see Kirsten E. Schulze, The
federacy 243
Free Aceh Movement (GAM): Anatomy of a Separatist Organization (Washington, DC: East-West
Center, 2004).
98. Edward Aspinall and Harold Crouch, ‘‘The Aceh Peace Process: Why It Failed,’’ Policy
Studies 1 (Washington, DC: East-West Center, 2003), p. 1.
99. On Freedom House’s seven-point scale for political rights—where 1 is the best possible score
and 7 the worst—Indonesia receives a 2. See Freedom in the World: Political Rights and Civil
Liberties: 2010 (New York: Freedom House, 2010), available at www.freedomhouse.org. For the
argument by two leading political scientists who specialize on Indonesia that Indonesia is close to
democratic consolidation, see R. William Liddle and Saiful Mujani, ‘‘Indonesian Democracy:
From Transition to Consolidation,’’ paper prepared for an International Conference on ‘‘Islam and
Democracy in Indonesia: Comparative Perspectives,’’ Center for Democracy, Toleration, and Reli-
gion, Columbia University, New York, April 2–3, 2009. The World Bank quotation is from Samuel
Clark and Blair Palmer, ‘‘Peaceful Pilhada, Dubious Democracy: Ache’s Post Conflict Elections
and Their Implications,’’ Indonesian Social Development Papers, no. 11 (Jakarta: World Bank, No-
vember 2008), p. vi. The word ‘‘dubious’’ comes from Clark and Palmer’s focus on the need to
improve the quality of election procedures and practices in order to deepen democracy. Indeed, in
the runup to the 2009 elections, while there were no signs of the resurgence of the civil war, there
was some use of intimidation and violence by participants in the electoral struggle.
244 crafting state-nations
helped the peace outcome. Three of the events have been well written about, but
in our judgment none were sufficient.
First, a tsunami killed upwards of 150,000 people in Aceh on December 26,
2004. Within a week, GAM and Susilo Bambang Yudhoyono, the newly elected
president, expressed interest not only in a ceasefire but also in negotiations. How-
ever, as a single causal exogenous explanation, the tsunami disaster is not satisfac-
tory. The same tsunami smashed neighboring Sri Lanka, with the opposite politi-
cal effect. The struggle over relief resources worsened. Eventually the Norwegian
peacekeeping mission was asked to leave so that the government could make one
more attempt to eliminate the Tamil Tigers (see chapter 5). Furthermore, it has
also now been documented that significant preparations for secret peace discus-
sions about Aceh had already begun at least a month before the tsunami.100
The second explanation draws on Dankwart Rustow’s classic ‘‘Transitions to
Democracy: Toward a Dynamic Model,’’ where he advances the argument that if
two opponents are engaged in an exhausting major political struggle, and one side
can defeat the other, victory by one side, rather than democratic concession or
democratic peace accords, may well be the outcome. For Rustow it is precisely
‘‘prolonged and inconclusive political struggle’’ that can be conducive to the
construction of new master frameworks of peace or democracy.101 Modern con-
flict settlement theory employs the analogous concept of ‘‘hurting stalemates.’’102
Specialists on the military dimension of the conflict in Aceh have made a strong
case that a stalemate had occurred by late 2003. The May 2002 military offensive
against GAM had indeed taken a great toll. GAM had been forced to retreat to the
mountains and had virtually ceased offensive actions. Nonetheless, even though
the commander of the Indonesian armed forces had vowed to root out GAM
completely, he later reluctantly acknowledged that they could not be defeated:
‘‘We cannot do what we hoped. . . . Two die but four take their place.’’103 However,
100. See the memoirs of a key Australian adviser to the GAM, Damien Kingsbury, Peace in
Aceh: A Personal Account of the Helsinki Peace Process (Jakarta: Equinox Publishing, 2006), pp. 15–
22; and Edward Aspinall, ‘‘The Helsinki Agreement: A More Promising Basis for Peace in Aceh?’’
Policy Studies 20 (Washington, DC: East-West Center, 2005), pp. 14–19.
101. See Dankwart Rustow, ‘‘Transitions to Democracy: Toward a Dynamic Model,’’ Compara-
tive Politics 2 (April 1970), p. 352.
102. See, for example, William Zartman, ‘‘The Timing of Peace Initiatives: Hurting Stalemates
and Ripe Moments,’’ Global Review of Ethnopolitics, no. 1 (2001), pp. 8–18. In fact, one of the
sections of Aspinall’s ‘‘Helsinki Agreement’’ monograph is called ‘‘Toward a Hurting Stalemate?’’
103. Statement of Armed Forces Chief Endriartono Sutarto, cited in Aspinall, ‘‘Helsinki Agree-
ment,’’ p. 21. True, Sutarto did not make this statement until 2005, but evidence suggests that this was
the situation by late 2003.
federacy 245
even if there were a complete stalemate, the theory is not necessarily a sufficient
explanation in itself. Stalemate does not necessarily produce peace if the long-
term goals of the two adversaries are still locked in a zero-sum relationship. GAM
could and did withdraw to the mountains with virtually all their leaders and
command structure intact. The military could and did continue to occupy all the
urban space of Aceh. The new equilibrium was simply a less hurting stalemate, a
possibly fragile ceasefire but certainly not yet peace.
The third explanation of the origins of the surprising peace was the election in
November 2003 of retired military general Susilo Bambang Yudhoyono (normally
referred to as ‘‘SBY’’) as president and Jusuf Kalla as vice president; both were
creative, hands-on advocates of a serious new peace process in Aceh.104 This was
certainly extremely helpful, and a new peace process began in Helsinki in January
2003. However, the emergence of these new leaders was not necessarily sufficient
for the achievement of peace. A strong case could be made that unless GAM had
radically and credibly renounced its goal of independence, even these new lead-
ers would not have been able to build and sustain a winning peace coalition
among the nationalist majority in the military and the parliament, whose collabo-
ration was necessary for the crafting of a sustainable peace. Indeed, the first round
of the Helsinki Peace talks, on January 27–30, 2005, yielded nothing but pessi-
mism. In fact, in Aspinall’s judgment, ‘‘On the eve of the second round of talks on
February 21–23, 2005, the gulf between the two sides seemed as wide as ever.
Collapse seemed a real possibility.’’105
However, soon after the second round began, a spokesperson of GAM an-
nounced at a press conference on February 23, 2005: ‘‘The demand for Indepen-
dence is no longer on the table. They are demanding self-government now and
the Indonesians [government representatives] understand this very clearly.’’106 As
Aspinall put it, ‘‘This was a shift of historic proportions. It was the first time that
GAM had ever indicated that it was prepared to accept anything less than inde-
pendence or a referendum [on independence]. As such, it was widely viewed as
major breakthrough, and it made all subsequent progress in the talks possible.’’107
104. On the contributions of this exceptional new SBY-Kalla team, the most detailed report of
the government side of the Helsinki negotiations, based on extensive interviews with SBY and Kalla
and all the key government negotiators, is Michael Morfit, ‘‘The Road To Helsinki: The Aceh
Agreement and Indonesia’s Democratic Development,’’ International Negotiation 12, no. 1 (2007),
pp. 111–143.
105. Aspinall, ‘‘Helsinki Agreement,’’ p. 27.
106. Associated Press, February 23, 2005, cited in Aspinall, ‘‘Helsinki Agreement,’’ p. 26.
107. Ibid.
246 crafting state-nations
not sleep that night. They spent the night looking up the Åland Islands, and then
Greenland, on the Internet. They may or may not have heard of the word fed-
eracy, but they increasingly began to feel that the Finnish–Åland Islands federacy
arrangement might produce a serious form of ‘‘self-government’’ for Aceh.111
Whatever the exact reasons for GAM’s withdrawal of their demand for inde-
pendence, the bargaining situation in Helsinki was no longer zero-sum. A non-
zero-sum game was now on the table. If GAM were able to negotiate for a federacy
of the Åland Islands type, they believed many of their goals concerning self-
government could be met. If the central government’s negotiators were able to get
such a federacy, their central goal of peacefully keeping Aceh inside of the unitary
state of Indonesia would be met. Such a mutually acceptable treaty, called the
‘‘Memorandum of Understanding between the Government of the Republic of
Indonesia and the Free Aceh Movement’’ (MoU), was signed on August 15, 2005,
and witnessed by Martti Ahtisaari.112
The major goal of the central government was achieved in the second para-
graph of the MoU Preamble: ‘‘The parties commit themselves to creating condi-
tions within which the government of the Acehnese people can be manifested
through a fair and democratic process within the unitary state and constitution of
the Republic of Indonesia’’ (emphasis added).
In addition to forgoing their independence claims, GAM agreed to demobilize
‘‘all of its 3000 military troops’’ and to undertake ‘‘the decommissioning of all
arms, ammunition and explosives held by the participants in GAM activities’’
under the supervision of the newly created EU-ASEAN–led ‘‘Aceh Monitoring
Mission,’’ no later than December 31, 2005.
The quid pro quo for GAM were new federacy prerogatives, many quite similar
to those agreed to in the Åland Islands–Finland federacy, that were to be the
principles of a new Law on the Governing of Aceh. Four of our five defining
characteristics of a federacy, and both of our facilitating features, are clearly
involved in the MoU.
Defining Characteristic 1: Federal-like division of state and federacy functions.
For GAM negotiators, some of the most important self-governing prerogatives
were spelled out in Article 1 of the MoU, which set out federal-type arrangements:
‘‘Aceh will exercise authority within all sectors of public affairs . . . except in the
111. Interviews of Alfred Stepan with Nur Djuli in Banda Aceh, Aceh, Indonesia, November 2,
2007, and Bireuen, Aceh, November 4, 2007.
112. A complete English copy of the MoU is available in Aspinall, ‘‘Helsinki Agreement,’’
pp. 75–84.
248 crafting state-nations
fields of foreign affairs, external defense, national security, monetary and fiscal
affairs, and justice and freedom of religion.’’113
Defining Characteristic 3: Existence of dispute resolution procedures. In most
policy areas where Aceh would not be self-governing because they involved state
powers, the MoU stipulated that policies would be crafted ‘‘in consultation with
and with the consent of ’’ the relevant authorities of Aceh. Specifically, Article
1.1.2(c) stipulates that ‘‘Decisions with regard to Aceh by the legislature of the Re-
public of Indonesia will be undertaken in consultation with and with the consent
of the legislature of Aceh,’’ and 1.1.2(d) that ‘‘Administrative measures undertaken
by the Government of Indonesia with regard to Aceh will be implemented in
consultation with and with the consent of the head of the Aceh administration.’’
In police and military affairs, an important distinction was made in the MoU
between ‘‘organic’’ and ‘‘non-organic’’ personnel. The Government of Indonesia,
subject to binding verification by the Aceh Monitoring Mission, agreed to ‘‘with-
draw all elements of non-organic military and non-organic police forces from
Aceh’’ (Article 4.5).114 Very importantly, ‘‘The appointment of the Chief of the
organic police forces and the prosecutors shall be approved by the head of the
Aceh administration’’ (Article 1.4.4).
Defining Characteristic 4: Reciprocal representation between the unitary state
and the federacy. Under the section called ‘‘Political Participation’’ (Article 1.2), it
is explicitly stated that ‘‘Full participation of all Acehnese people in local and
national elections will be guaranteed in accordance with the Constitution of the
Republic of Indonesia.’’ The MoU is silent on the question of the unitary state
having a representative in the federacy; however, it does speak of central state
functions being carried out by a variety of state representatives.
Defining Characteristic 5: The federacy is part of an internationally recognized
independent state. Indonesia is a member of the United Nations, but Aceh is not.
Facilitating Characteristic 1: Role of international guarantors in the founding of
the federacy. The chair of the Helsinki peace process that created the federacy was
not an Indonesian but rather Martti Ahtisaari, the former president of Finland,
who was also a major UN negotiator in Bosnia. Furthermore, the MoU explicitly
states in Article 5.1 that ‘‘An Aceh Monitoring Mission (AMM) will be established
113. Indonesia’s general decentralization law of 1999 had given extensive new rights to prov-
inces, but GAM negotiators evidently saw Article 1 of the MoU as more significant and more binding
on Jarkarta.
114. ‘‘Non-organic’’ personnel meant any extra central government troops that had been sent in
to bolster the numbers and capacity of the regular ‘‘organic forces.’’ This distinction implied a
significant reduction of the number of Indonesian government security forces in Ache.
federacy 249
by the European Union and ASEAN contributing countries with the mandate to
monitor the implementation of the commitments taken by the parties in this
Memorandum of Understanding.’’ In cases of a dispute between the government
of Indonesia and GAM, Article 6.1 stipulates that ‘‘The Head of the Monitoring
Mission will make a ruling which will be binding on the parties.’’ Some of the key
issues that the AMM monitored were the disarmament of GAM, the exit of the
‘‘non-organic’’ Indonesian military forces from Aceh, and the supervision of free
elections in Aceh—all three of these issues were crucial to the successful setting
up of the federacy in Aceh.
Facilitating Characteristic 2: Role of the federacy in international treaties signed
by the center. The MoU may be more explicit about this than any other federacy.
Article 1.1.2(b) asserts that ‘‘International agreements entered into by the govern-
ment of Indonesia which relate to matters of special interest to Aceh will be
entered into consultation with and with the consent of the legislature of Aceh.’’
While they are not defining conditions, some other aspects of the MoU helped
end the civil war and facilitate the construction of the federacy and the deepening
of democracy. In one of the most contested and important prerogatives, the MoU
stated that, ‘‘understanding the aspirations of Acehnese people for local political
parties, the Government of Indonesia will create . . . the political and legal
conditions for the establishment of local political parties’’ (Article 1.2.1). This was
crucial for GAM, because Indonesia’s parliament, fearing regional fragmentation,
had passed an exceptionally restrictive law mandating that all political parties
must have a recognized party committee in at least 60% of Indonesia’s provinces
and in 50% of the districts in those provinces. If a party does not meet these
requirements, it is not recognized as a legal body.115 Such a law would, of course,
have banned the most important ethno-regional parties in Canada, Spain, Bel-
gium, and India, all of which we have discussed in this book. Without this special
prerogative, GAM, being exclusively Aceh-based, would not have been able to
run for the upcoming elections as a political party. With this special prerogative,
however, GAM, running as independents but clearly representing GAM, was able
to enter democratic political competition for the governorship and for all district
heads and mayors in Aceh in the December 2006 elections. To ensure fair elec-
tions, there was an EU Election Monitoring team present. To the surprise of many
people, the formerly Aceh-based head of intelligence for GAM, Irwandi Yusuf,
won the governorship. Also to the surprise of many, the president of Indonesia
immediately sent his warmest congratulations. After the MoU (and the tsunami),
the governor’s office was worth controlling. According to the World Bank, the
budget of Aceh in 2008 would be around 1.5 billion dollars, seven times that of the
1999 figure. GAM also won about a third of the mayorships in Aceh.116 In addition
to these federacy-like prerogatives, a number of provisions ending the civil war
were offered to GAM, such as amnesty, release of all political prisoners, and
reintegration funds to be administered by GAM.117
Four months after the Helsinki Agreement was signed and the deadline for
disarmament arrived, one of Aceh’s longest and most respected observers, Sid-
ney Jones, released a report declaring that ‘‘the Aceh peace process is work-
ing beyond all expectations. Guerrillas of the Free Aceh Movement, GAM, have
turned in the required number of weapons. The Indonesian military, TNI, has
withdrawn troops on schedule. The threat of militia violence has not mate-
rialized. Amnestied prisoners have returned home without incident. The Inter-
national Aceh Monitoring Mission, AMM, led by the European Union’s Peter
Firth, has quickly and professionally resolved the few violent incidents between
GAM and the TNI. . . . The peace process has active support of the Indonesian
government.’’118
Recently, two events have happened that some interpret as implying GAM
support for harsh shari’a and violence. Neither event does. In October 2009, four
years after the Helsinki agreement, the legislature of Aceh passed a law supporting
death by stoning in cases of adultery. Two things should be noted. One, this law
was passed in the last days of an outgoing legislature that contained no GAM
members. The legislature had been based on elections that occurred before GAM
could compete as a party in elections. After the law was passed, the governor of
Aceh, Irwandi Yusuf, who was the former head of GAM intelligence for Aceh,
116. See the article by Edward Aspinall, ‘‘Guerrillas in Power,’’ in the special issue edited by
Edward Aspinall, ‘‘Aceh: Two Years of Peace,’’ in Inside Indonesia, no. 90 (October–December
2007). Stepan visited a former GAM exile, Dr. Nurdin, who was elected district head of Bireuen, one
of the most militarily contested war zones of Aceh. Nurdin said that the process was working
reasonably well despite the local history of great conflict. He had the flags of Aceh and Indonesia
and the standard photograph of the president of Indonesia that one sees in every public office in
Indonesia above his desk. The data on Aceh’s 2008 budget were given to Stepan by the World Bank
office in Aceh.
117. MoU, Section 3, ‘‘Amnesty and Reintegration into Society.’’
118. See International Crisis Group, ‘‘Aceh: So Far, So Good,’’ Asia Briefing, no. 44, December
13, 2005, p. 1. GAM was both proactive and active in the peace process; for example, one of the most
senior former GAM commanders, Darwis Jeunieb, turned over to the police a former GAM soldier
who was still trying to enforce GAM’s now renounced ‘‘intelligence tax,’’ and GAM leaders were in
charge of the compensation program for the demobilized GAM troops, pp. 4, 6.
Stepan interviewed Sidney Jones in Jakarta on November 4, 2007. For additional corroborative
accounts of the peace dimensions two years after Helsinki, see the special issue edited by Edward
Aspinall, ‘‘Aceh: Two Years of Peace.’’ Stepan interviewed Aspinall in Aceh on November 6, 2007.
federacy 251
criticized the law and said that he would never implement it.119 The second major
event was the discovery that a terrorist base camp was being constructed in Aceh
in April 2010. In their assessment on the event, the International Crisis Group
asserted, ‘‘Former rebels from [GAM] were not involved in any significant way.’’120
They also note that the information about the base was given to the police by an
Acehinese.
In a survey conducted in 2006 among 1,075 former GAM militants by Colum-
bia University professor Macartan Humphreys, the former GAM militants were
asked to indicate their degree of satisfaction or dissatisfaction ‘‘with MoU imple-
mentation.’’ ‘‘Satisfied’’ former GAM militants outnumbered ‘‘dissatisfied’’ former
GAM militants nine to one: 46% were ‘‘very satisfied,’’ 44% ‘‘satisfied,’’ 7% ‘‘not
satisfied,’’ 2% ‘‘very dissatisfied,’’ with only 1% refusing to answer.121 The prospect
of a federacy—and of peace—was an acceptable alternative to the former fighters
for independence.
This section would not be balanced, however, if we did not point out how Aceh
falls short of the ideal-type federacy we discussed at the beginning of this chapter.
Aceh is the only case of federacy-type arrangements being constructed by two
forces actively engaged in civil war. In the case of Aceh, this contributed to one
major limitation. The MoU was a peace agreement signed by the government of
Indonesia and GAM. But in the new democracy of Indonesia, MoU could not
become a law, much less be constitutionally embedded, until the parliament
ratified it. The sitting parliament, without GAM representatives—because GAM,
as a revolutionary force turned provincial political party, would participate in
parliamentary elections for the first time in 2009 as Partai Aceh—passed a Law on
the Governing of Aceh (LOGA) that incorporated many but by no means all parts
of the MoU.122 The most important watered-down provision was that instead of
LOGA stipulating that laws, treaties, or administrative measures affecting Aceh
must be with the ‘‘consultation and consent’’ of the relevant Aceh authorities, only
the word consultation was included, not consent. Also, in our ideal type of
federacy the arrangements should be constitutionally embedded. However, to
GAM’s great disappointment, LOGA is a law that can be amended by an ordinary
119. In a conversation with Alfred Stepan on May 19, 2010, Sidney Jones, the widely respected
project director for the International Crisis Group in Jakarta, told Stepan that she doubted whether
the law would ever be implemented in Aceh. We hope the law will be formally abrogated.
120. International Crisis Group, ‘‘Indonesia: Jihadi Surprise in Aceh,’’ Asia Report, no. 189,
April 20, 2010.
121. Unpublished results provided to Alfred Stepan by Macartan Humphreys on March 6, 2009.
122. See, President, Republic of Indonesia, Law of the Republic of Indonesia, Number 11 of the
year 2006 Regarding Governing of Aceh.
252 crafting state-nations
majority in the lower house of parliament.123 At the very least, we have to acknowl-
edge that our second defining condition of a federacy, constitutional embedded-
ness, is not fully met.124
Nonetheless, we think that we have produced strong evidence indicating that
the surprising emergence of social peace and political incorporation in Aceh
would not have been possible without the idea of federacy and without the utiliza-
tion of many important federacy-type arrangements in Indonesia. In an October
2009 interview with Stepan, Vice President Kalla stressed the helpfulness of feder-
acy for the peace process.
In future studies of federacy, more consideration of the particular asymmetrical
bargaining powers of armed separatists versus the government in Time 1 (during
negotiations) and in Time 2 (after the separatists’ disarmament and participation in
elections) should be analyzed. During Time 1, both the government and the sepa-
ratists may have strong incentives and capacities to make federacy arrangements.
However, during Time 2, if the former separatists want to return to armed violence,
their disarmament, demobilization, and the public disclosure of members in hid-
ing will have greatly decreased their capacity to launch another rebellion. Also, be-
cause many of their most effective fighters may now be in important public
positions of power, they will have less incentive to resist. Asymmetrically, the
government’s capacity and incentive to resist a new rebellion would in no way be
diminished; in fact, the government’s relative capacity to resist would be increased.
Building on (but at times going beyond) the cases discussed in this chapter, let us
offer some reflections as to why federacies might emerge, and the conditions that
123. Nur Djuli, the GAM negotiator at Helsinki who told Stepan about his conversation with
the former President of Finland, Martti Ahtisaari, about the ‘‘self-governing’’ qualities of federacy in
the Åland Islands, felt particularly betrayed by the changes by parliament and constructed and
circulated a ‘‘matrix’’ illustrating how some key guarantees of autonomy were less strong in the
LOGA than in the MoU. Indeed, the danger of the parliament attempting to erode MoU further was
evident in January 2008, when, led by former President Megawati Sukarnoputri’s coalition building
efforts among the opponents of MoU, the parliament passed a law recommending that in a future
reorganization of Indonesia’s administrative units, some units that are now a part of Aceh should be
transferred elsewhere. This would not only be a violation of MoU but also of LOGA. The bill never
went into effect because President Yudhoyono did not sign it. See ‘‘Indonesia: Pre-Election Anxi-
eties in Aceh,’’ International Crisis Group, Asia Briefing, no. 81, September 9, 2008, pp. 6–8.
124. LOGA, by what constitutional theorists call embedded political ‘‘convention,’’ might have
become de facto constitutionally embedded, and therefore requiring exceptional majorities to
amend, but the Megawati initiative would indicate that such a convention was lacking for her. For
the classic treatment of political constraints and conventions producing unwritten forms of constitu-
tional embeddness, see Marshall, Constitutional Conventions.
federacy 253
might support or hinder their relative success, from the perspective of irredentas
and identities.
The crafting of a federacy is less complicated if there are no irredentist neigh-
bors. Denmark had no neighbors with irredentist ambitions toward the Faroe
Islands or Greenland. Some secessionists in the Azores were interested in being
an ‘‘associated state’’ of the United States, but as soon as the Social Democrats and
their allies had clearly secured control, the secessionists no longer aspired to this
exit option; in any case, the United States never explicitly articulated any version
of irredentism comparable to that articulated by Austria toward South Tyrol or
Yugoslavia toward much of Friuli-Venezia Giulia and Trieste. In the case of Aceh,
no neighbor articulated irredentist claims, and all of Indonesia’s neighbors partici-
pated as monitors of the Helsinki Peace Agreement.
However, this does not necessarily mean that the existence of an irredentist
power reduces the likelihood of a federacy being attempted. Quite the opposite. If
a unitary state is faced with a secessionist territory but does not want to use
coercion, grant independence, or allow its annexation, the existence of an irre-
dentist neighbor could well be an incentive to adopt a federacy as a ‘‘holding
together’’ strategy, because a de facto federacy may be the best available option.
Furthermore, federacy arrangements, particularly if an irredentist neighbor is
part of a network of democratic powers, much more than a unitary state arrange-
ment or even asymmetrical federal arrangements, can allow space for foreign
states or international organizations to be involved in negotiating, and eventually
endorsing or even guaranteeing, arrangements to protect the rights of members of
a minority linguistic community. Witness the De Gasperi–Gruber Accord be-
tween the foreign ministers of Austria and Italy and the later appeal, building on
this accord, by Austria to the United Nations that led to the 1972 and 1992 exten-
sions of linguistic, cultural, and political rights in the South Tyrol statute of
autonomy. Recently, the North Tyrol province of Austria, and the South Tyrol
province of Italy, wanted to engage in bilateral negotiations to create a Euro-
region. The governments of Italy and Austria discouraged this, because both felt
that it might endanger the carefully crafted and negotiated cultural autonomy
statutes and violate the preeminent role of the sovereign states in foreign policy
affairs.125
The settlement of the Åland controversy, with the award by the League of
Nations of the islands to Finland and not to Sweden, was facilitated by the
125. See Bruno Luverà, ‘‘L’Euregio tirolese. Tra regionalismo transfrontaliero e micronaziona-
lismo di confine,’’ in Altre Italie, ed. Nevola, pp. 19–33.
254 crafting state-nations
conclusion
We live in a world of both states and nations. Federacies are a response to the
problems of ‘‘holding together’’ a unitary state while also responding to the claims
of the minority nation principle within that state. Federacies can work, but only
if there is minimum loyalty to the state and if the principle of national self-
determination is not exploited by other states. In the case of an area that has been
considered an ‘‘irredenta’’ by a neighboring state, this means support of the fed-
256 crafting state-nations
eracy by that potentially irredentist power. The legitimacy of the ‘‘voice’’ and the
‘‘exit’’ of the federacy population, in addition to the ability of the unitary state to
control its own elites, is often a result of the world—especially the regional—
system of states within which the possible federacy would exist. What is often
called the international dimension of politics is of course more often an interstate
dimension. As we write this, we think of the interstate system that would have
affected a potential federacy in South Ossetia in 1991 in contrast to that facing
South Tyrol in 1945. Ultimately, we are living in a world of states, united or dis-
united, and not of nations—but also of nations disturbing the status quo of states.
chapter eight
The purpose of this chapter is to explore the question: How appropriate or in-
appropriate is U.S.-style federalism in robust multinational societies if our goals
are democracy, reasonably inclusive social welfare policies, and relative political
tranquility? We ask this question because, even though we argued in chapter 1 that
the United States was multicultural but not multinational, many readers may
have reservations with both the ideal types we have examined so far, nation-state
and state-nation. Why then, they may ask, not just follow the U.S. model, espe-
cially since we have stressed the critically helpful role of federalism?
For many theorists and political leaders, the U.S. model of federalism is held
up as not only the first, but also the most authentic—indeed, the best—model
of federalism. Thus, one of the most prestigious political scientists of federalism
in the English-speaking world, William H. Riker argues in many of his works that
U.S. federalism is ‘‘the model to which all others aspire.’’1 In a very influen-
tial early analysis of Indian federalism Kenneth C. Wheare termed India’s model
‘‘quasi-federalism,’’ mainly because of its deviations from the U.S. model.2 Even
in modern Spain, some influential critics do not consider Spain’s current model
‘‘federal,’’ because it was not, as in the United States, created by autonomous
political units that came together and decided to pool some of their sovereignty
1. William H. Riker, Federalism: Origin, Operation, Significance (Boston: Little Brown, 1964).
See also his influential The Development of American Federalism (Boston: Kluwer Academic Pub-
lishers, 1987) and ‘‘Federalism,’’ in Handbook of Political Science, ed. Fred Greenstein and Nel-
son W. Polsby (Reading, MA: Addison-Wesley, 1975), vol. 5, pp. 93–172.
2. Kenneth C. Wheare, Federal Government (Oxford: Oxford University Press, 1963), pp. 26–28.
258 crafting state-nations
to create a federal system. Indeed, on more than one occasion during our re-
search, it was clear to us in conversations with important politicians in unitary
countries such as Sri Lanka, Indonesia, and the Philippines, that one of the
reasons they had difficulty considering a federal option is that, considering the
U.S. model to be the most authentic, they thought that they would have to
dissolve the unitary state and then let those units who wanted to join have a
constituent assembly in order to construct, as in Philadelphia in 1787, a federal
constitution. Many of the leaders of nationalist armies in Burma who have been
struggling for independence, when they got together to draft a constitution in
which they would possibly agree to remain within Burma, looked to the U.S.
model as the one that would be the be the most appropriate, that would most
protect their rights.
Given all the above arguments, we think it is imperative for us to try to answer a
fundamental question. Is the U.S. federal model particularly useful for demo-
cratic governance in robustly multinational societies, is it neutral in its impact, or
could it be particularly harmful?
We believe that we can get some leverage on these issues by carrying out a
simple but potentially powerful exercise revolving around two interrelated sub-
questions we will ask. First, what are the key features of the specific type of
federalism in the United States? How do they relate to the rest of the U.S. Consti-
tution? How do they compare with the features found in the other seven long-
standing OECD advanced-economy federal systems? These countries are Ger-
many, Austria, Belgium, Spain, Switzerland, Canada, and Australia. Note that
four of these countries, Belgium, Canada, Switzerland, and Spain, are close to the
state-nation ideal type. Second, what are some of the implications for governance
in a robust multinational society, if the entire U.S. package of distinctive federal
features were adopted?
We realize that this is a thought exercise. No country would probably adopt the
entire package. However, carrying out such an exercise will generate great ana-
lytic (and political) opportunities. It will enable us to analyze many aspects of U.S.
federalism comparatively, both in its component parts and as an interrelated
system. In the preface of this book, we argued that one of our purposes was to
expand our imaginations concerning possible new and constructive ways of man-
aging multinational polities. We hope our concept of the state-nation has ex-
panded political imaginations. Another way of doing this is to carry out a thought
exercise like the one we will do now, one that asks us to systematically ‘‘re-
imagine’’ what a well-known, classic model, such as U.S. federalism, may or may
not actually imply.
the u.s. federal model and multinational societies 259
1. The upper chamber is extremely malapportioned. Each state in the U.S. fed-
eration, no matter how large or small its population, receives the same number of
seats in the Senate. This is, of course, a massive violation of the democratic
principle of ‘‘one person, one vote.’’ Article 5 of the U.S. Constitution stipulates
that ‘‘the Senate of the United States shall be composed of two Senators from each
State.’’ This means that, in terms of producing a U.S. senator, a vote in Wyoming,
with its population in 2009 of only 544,270 people, had the approximate value of
sixty-eight votes in California, with its population of 36,961,664.3 Furthermore,
Article 5, one of the least democratic articles in the Constitution, is the only part of
the constitution that was framed to be impossible to amend without complete
unanimity in the Senate. The Constitution expressly mandates that ‘‘no state,
without its consent, shall be deprived of its equal suffrage in the Senate.’’4 In
Robert A. Dahl’s judgment, ‘‘The likelihood of reducing the extreme inequality of
representation in the Senate is virtually zero.’’5
In comparison with the other members in our set of the eight longstanding
federal democracies in advanced economies, as shown in table 8.1, the United
States has the most extreme Gini index of malapportionment.
2. The upper chamber has major and unique constitutional powers. This ex-
tremely malapportioned chamber, the U.S. Senate, has an absolute veto on all
legislation (Article 1, section 7). The Senate also has many exclusive and critical
prerogatives denied to the very well apportioned House of Representatives, such as
the veto power over all presidential nominations for major judicial positions,
cabinet members, and heads of major government agencies as well as the ratifica-
tion of treaties (Article 2, section 2). Also, the ‘‘Senate shall have the sole power to
try all impeachments’’ (Article 1, section 3).
In comparative terms of our set of eight countries, the upper chambers in
Austria and Belgium do not have an absolute veto. Canada’s upper chamber de
jure has an absolute veto but, as an appointed body, de facto would not dare use it.
3. Population estimates come from the U.S. Census Bureau. See http:/ /quickfacts.census.gov/
qfd/states/56000.html for Wyoming and http:/ /eadiv.state.wy.us/pop/st-09est.htm for a fifty-state
comparison.
4. U.S. Constitution, Article 5.
5. See the excellent and sobering book, Robert A. Dahl, How Democratic Is the American
Constitution? (New Haven: Yale University Press, 2001), p. 154. For some of the negative effects of
this malapportionment in the Senate on social policies aimed at reducing inequality in the United
States, see pp. 13–15, 46–54, and 144–145.
260 crafting state-nations
table 8.1
Overrepresentation in the Upper Houses of Eight Longstanding OECD
Federal Democracies and India
Ratio of best-represented to
Gini index of worst-represented federal Percentage of seats of best-
malapportionment unit (by population) represented decile
Nation Index Nation Ratio Nation %
United States .49 United States 66/1 United States 39.7
Switzerland .45 Switzerland 40/1 Switzerland 38.4
Australia .36 Canada 21/1 Canada 33.4
Canada .34 Australia 13/1 Australia 28.7
Germany .32 Germany 13/1 Germany 24.0
Spain .31 India 11/1 Spain 23.7
India .10 Spain 10/1 India 15.4
Austria .05 Belgium 2/1 Austria 11.9
Belgium .02 Austria 1.5/1 Belgium 10.8
Sources: Data come from Whitaker’s Almanack (London: I. Whitaker, 1997); The Europa World Year Book 1995
(London: Europa Publications, 1995); and Daniel Elazar et al., Federal Systems of the World (Harlow, U.K.: Long-
man, 1994). For the constitutional provisions on second chambers, see S. E. Finer, Vernon Bogdanor, and Ber-
nard Rudden, Comparing Constitutions (Oxford: Oxford University Press, 1995), and A. P. Blaustein and G. H.
Flanz, eds., Constitutions of the Countries of the World (Dobbs Ferry, NY: Oceana Publications, 1991–).
Note: The inequality (malapportionment) index is computed as (1/2) 兺 |si – vi| where si = % of seats and
vi = % of population. This calculation follows the formula proposed by David Samuels and Richard Snyder
in ‘‘The Value of a Vote: Malapportionment in Comparative Perspective,’’ British Journal of Political Science
(October 2001), p. 655.
In Germany, the upper chamber has an absolute veto for only about 60% of
legislation. In Spain, the upper chamber has a veto only on issues that directly re-
late to the constitutionally embedded prerogatives of the member units of the fed-
eration. Only the three classic ‘‘coming together federations,’’ the United States,
Australia, and Switzerland, have an absolute veto on all legislation.
The upper chamber, by itself, seldom blocks legislation in Switzerland, occa-
sionally does so in Australia, and frequently does so in the United States. None of
the other upper chambers in the set has as many unique prerogatives as does the
U.S. Senate.6
3. U.S. federalism is symmetrical, not asymmetrical. The framers negotiated a
Constitution in which every state has the same rights and obligations, that is to say,
U.S. federalism is symmetrical, not asymmetrical. In an asymmetrical federation,
some federal members could constitutionally be given special prerogatives con-
cerning some cultural features. This symmetrical feature of U.S. federal arrange-
6. For details and documentation on the veto powers of the second chambers in all advanced
economy OECD countries, see Stepan, ‘‘Electorally Generated Veto Players.’’
the u.s. federal model and multinational societies 261
ments is not unusual in aggregate comparative terms, because more than half
of our set (Germany, Austria, Australia, and Switzerland, as well as the United
States) are symmetrical. But, if we disaggregate our set so as to identify those
polities that have a territorially based robust multinational dimension, we notice
that all of them (India, Spain, Belgium, and even Canada) are asymmetrical.
As we saw in our discussion of ‘‘policy grammars’’ and have seen in the rest of this
book, whether a constitution is symmetrical or asymmetrical has important conse-
quences for what can be done, or especially for what cannot be done, to manage
politics in a polity with multinational dimensions. To return to India and Mizoram,
a key part of the agreement to end the civil war was to constitutionally grant two
special prerogatives to the state of Mizoram. Non-Mizos are prohibited from buy-
ing property in Mizoram, and non-Mizos cannot run in a local election.
4. ‘‘Residual powers’’—not to the union but to the individual states. In the
demarcation of union and state tasks in the U.S. Constitution, the states of the
federation were assigned many key tasks. On top of this, any tasks that were not
explicitly assigned to the federal center were assumed not to be federal, or joint,
but under the authority of the states; that is, ‘‘residual powers’’ were granted to the
states.7 In Spain, residual powers belong to the center; in Germany most powers
are de facto concurrent.
5. Senate and states must both vote approvals of any amendment by super-
majority. The above four features, taken together, are not only (as we shall see)
highly unusual in the world of developed country democratic federal systems but
are also embedded in a constitution that is one of the hardest, if not the hardest, in
the world to change. As John Elster has argued, the most difficult rules to change
are those that most favor actors who are charged with amending them. In the case
of the U.S. Constitution, exceptional majorities by the two political actors most fa-
vored by the Constitution, the Senate and the states, are required for any amend-
ment. An amendment must be approved by two-thirds of the state legislatures or of
both houses and then ratified by three-fourths of the states, either by special
conventions or by their legislatures (Article 5). On Lutz’s index for assessing the
relative difficulty of the amendment process, the United States is rated as having
the most difficult amendment process, not only of all of the countries in our eight
country set but also of the thirty-two countries calculated by Lutz.8
7. This was inserted into the U.S. Constitution under the 10th Amendment, which states, ‘‘The
powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are
reserved to the states respectively, or to the people.’’
8. See Donald S. Lutz, ‘‘Toward a Theory of Constitutional Amendment,’’ American Political
Science Review 88, no. 2 (June 1994), pp. 355–370. In Lutz’s rankings, the higher the index number,
262 crafting state-nations
Two other features of the U.S. political system are not about federalism per se
but about the separation of powers, within which U.S. federalism is embedded.
6. A supreme court with strong and extensive judicial review capacity. Even if
both houses and the president approve a bill to extend the powers of the federal
government via new legislation that they consider does not require an amend-
ment, this legislation can be declared unconstitutional by the Supreme Court, as
the Supreme Court has done in many cases.9
7. The U.S. presidential system entails a directly elected, ‘‘unsharable’’ executive
with a fixed term and strong veto powers. In a parliamentary system, executive
power is often ‘‘shared’’ by a coalition of parties. In the U.S. presidential system
however, the incumbent is necessarily one person, so the office is normally re-
ferred to as an ‘‘indivisible good’’ that is unsharable. The incumbent does not have
an absolute veto but is considered by Tsbellis to be a ‘‘veto player’’ because a
presidential veto can only be overridden by exceptional majorities in two houses
(Article 1, section 7 of the U.S. Constitution).10
the greater the degree of difficulty in amending the constitution. The United States scores highest
with a score of 5.1, whereas Australia, Portugal, Canada, and New Zealand all score less than 1.0.
India is 1.8.
9. In an analysis of all cases decided up to June 28, 2002, by the U.S. Supreme Court, the
Government Printing Office found that 158 acts of Congress had been declared unconstitutional.
See www.gpoaccess.gov/constitution/pdf2002/046.pdf.
10. See G. Tsbellis, Veto Players: How Political Institutions Work (Princeton: Princeton Univer-
sity Press, 2002); and G. Tsbellis, ‘‘Decision Making in Political Systems: Veto Players in Presiden-
tialism, Parliamentarism, Multicameralism, and Multipartyism,’’ British Journal of Political Science
25 (1995), pp. 289–325.
11. Even the unitary state of the United Kingdom has responded to the diversity of cultural
identities by a devolutionary process creating national legislatures: the Scottish Parliament (1999–),
the Welsh National Assembly (1998–), and the Northern Ireland Assembly (1998–). England itself
has no legislature.
the u.s. federal model and multinational societies 263
The reason why polities that have a multinational dimension to their society
find the asymmetrical formula politically useful is that it allows them, without
violating individual rights or the constitution, to give some special recognition to
some important cultural feature such as the language of a territorially concen-
trated minority nationality that also forms a majority in at least one of the units of
the federation. This is often seen by most designers of the federation as a legiti-
mate part of the strategy for maintaining unity amidst diversity, for nurturing
multiple but complementary identities.
Most analysts of the Spanish transition to democracy in the mid-1970s believe
that the successful transition was facilitated not only by federalism but also by the
ability, within an asymmetrical constitution, to recognize some historic preroga-
tives of the role of the Catalan language in Catalonia and tax privileges in the
Basque Country as well as recognition of the Basque language. In Canada, in
1867, the asymmetrical formula allowed a special place in Quebec for the Napo-
leonic Code as well as for the French language. Asymmetrical federalism in
Belgium granted special status to Wallonia and Flanders, while also according an
innovative ‘‘personal federalism’’ to Brussels. Personal federalism, which is not
found in Wallonia or Flanders, allows those citizens who self-identify as individ-
uals as members of different linguistic or religious communities to vote in separate
cultural constituencies for boards that supervise social services, such as hospitals
or schools.12 Similarly, India’s asymmetrical federal constitution would allow a
special status for Kashmir should peace ever be achieved. It also allowed the
central government to negotiate successfully an end to the decades-long war of
independence with the tribal Mizos by granting them a state with special rights
that would have been impossible in a symmetrical constitution. These special
Mizo rights included the right of their state assemblies to preclude non-Mizos
from buying property without their permission and the prohibition of non-Mizos
from voting in local, basically tribal elections. These special arrangements have
contributed, as we have seen, not only to the end of the civil war but also to high
degree of pride among Mizos in being Indian, even though less than 5% of the
Mizos speak Hindi or are Hindus.13
If we examine the entire universe of the eleven countries that have been contin-
uous federal democracies since 1988, we note that all of the polities that have a
territorially based multinational dimension to their societies are asymmetrical,
12. William Swenden, ‘‘Federalism and Second Chambers: Regional Representation in Parlia-
mentary Federalism,’’ D. Phil in the Department of Politics, University of Oxford, 2000.
13. See Jammu and Kashmir; Assembly Election 2002: Findings of a Post-Poll Survey by Lokniti,
Delhi, February 2003.
264 crafting state-nations
Constitution
Symmetrical Asymmetrical
Austria
Germany
Australia
Not Robustly
United States
Multinational
Argentina
Brazil
Switzerland (state-nation)
India
Robustly Belgium
Multinational Canada
Spain
whereas all of the others—many of which, like Switzerland, are deeply multi-
cultural but not multinational—have been able to function with symmetrical
federal systems.
The fundamental thing we want to stress is that a U.S.-style symmetrical consti-
tutional model would have made almost all of the special arrangements that we
have discussed for managing democracy in the territorially based multinational
contexts such as those found in Spain, Canada, India, and Belgium unconstitu-
tional and therefore politically impossible, in a democracy (see figure 8.1).
2. The weakness of coalitional incentives in the U.S. presidential system. Let us
now discuss the presidential, as opposed to the parliamentary, component of the
U.S. federal model and its theoretical and practical implications for democracy in
a robust multinational society.
For our purposes, the fundamental feature of presidentialism is that a directly
elected president, with a fixed term, is an ‘‘indivisible good,’’ that is, it cannot be
shared. One consequence of this is that if there is more than one nation in the
state, the president can only come from one of the nations. Also, such a president
can only be democratically removed by an impeachment, no matter what new
intra-nationalist conflicts or other problems might emerge in the short term.
the u.s. federal model and multinational societies 265
guarantee continued stability, and it does give the minority nationalities veto—
one might even say blackmail—potential. However, it does give the polity extra
‘‘degrees of freedom’’ to manage multinationalism via coalitions, degrees of free-
dom that would not be as broad if a polity selects, as part its constitutional formula,
a U.S.-style presidency.
3. The potentially severe political problem of every full federal unit having the
same number of seats (regardless of population) in the upper chamber. Is the
malapportion and inordinate power of the upper house in the U.S. federal model
less or more problematic in a robust multinational federal polity than it is in a
more homogeneous polity?
We think there could be demographic contexts where the U.S. formula for the
upper chamber would be potentially quite problematic. Let us take the hypotheti-
cal case of a potentially democratizing polity in a multinational society. Let us
assume that our hypothetical polity is territorially divided among segmented na-
tionality groups. Let us assume ten such groups, each of which has a state named
after the dominant nationality in that state.14
If state ‘‘A’’ contained the largest nationality in the polity and constituted 50%
of the total population in the polity but was only a majority in one state, state ‘‘A’’
would receive only 10% of the seats in the extremely powerful, U.S.-style upper
chamber we will call the Senate. To further complicate the scenario, let us make
the not impossible assumption that there is in state ‘‘B’’ a territorially concen-
trated, quite distinctive linguistic and religious nationality constituting only 1% of
the total population of the polity. Using the U.S. senatorial formula, state ‘‘A’’ and
state ‘‘B’’ would each receive the same number of seats in the Senate. This would
give nationality group ‘‘B’’ fifty times greater per capita representation in the
most powerful house in our hypothetical democracy than nationality group ‘‘A.’’
Clearly, this is not an auspicious recipe for a successful democratic transition.
Let us shift to a real country, where the situation is actually worse than the very
bad scenario just explored. In June 2001, Stepan participated in a meeting with
members of the Burmese democratic opposition, drawn heavily from political and
military leaders of nationality groups and armies.15 Most of the leaders had fought
long, but still unsuccessful, wars of independence. However, at the meeting the
guerrilla nationalist leaders wanted to discuss a draft constitution they had drawn
14. India is not a perfect example, but it is useful to remember that many of the current states in
India, such as Tamil Nadu, Kerala, Mizoram, Kashmir, and Bengal, are composed of a dominant
linguistic and ethnic group, some of whom have sometimes seen themselves as a nation.
15. This is reported in more detail in Andrew Reynolds, Alfred Stepan, Zaw Oo, and Stephen Le-
vine, ‘‘How Burma Could Democratize,’’ Journal of Democracy 12, no. 4 (October 2001), pp. 94–108.
the u.s. federal model and multinational societies 267
table 8.2
Malapportionment in Burma’s Proposed Upper Chamber
% of total Malapportionment
State Population population % seats index
‘‘Ministerial Burma’’ 40,647,593 75.93 12.5 63.43
Kayah 246,000 0.46 12.5 12.04
Chin 458,000 0.86 12.5 11.64
Kachin 1,202,000 2.25 12.5 10.25
Kayin 1,403,000 2.62 12.5 9.88
Mon 2,337,000 4.37 12.5 8.13
Rakhine 2,610,000 4.88 12.5 7.62
Shan 4,629,000 8.65 12.5 3.85
Total 53,532,593 100 100 63.42
Source: Population data for Burma, see www.myanmar.com.
up under which they would agree to stay within Burma. They were very attracted
by the U.S. federal model because it seemed to them to protect the interests of
their various states. Their draft constitution enshrined three key principles of U.S.
federalism. First, every state would receive an equal number of seats in the upper
house. Second, the upper house was made a bit stronger than the lower house, in
that all important executive positions were subject to the approval of the upper
chamber but not the lower chamber. Third, the draft constitution was symmetri-
cally federal.
In the guerrillas nationalist’s proposal, there would be eight states, each com-
posed of the territory of a major nationality. The U.S.-type decision rule meant
that two states, Kayah and Chin, each with less than 1% of Burma’s total popula-
tion, would receive the identical number of seats in the upper chamber as the
state known as Ministerial Burma, which has 76% of Burma’s population (see
table 8.2).
In terms of malapportionment, this proposal would have made Burma by far
the most malapportioned federal polity of the entire universe of the eleven federal
democracies in the world even if we add the non-OECD countries of India,
Brazil, and Argentina to our original eight-country advanced democracy com-
parison set (see table 8.3).
Over 90% of the Burmese officer corps is ethnic Burmese and from the state
called Ministerial Burma, which would be massively underrepresented by the
application of the U.S. formula. With 76% of the total population of Burma, the
state of Ministerial Burma would only receive 12% of the seats in the potentially
powerful upper chamber. This is the same percentage of seats as Kayah which has
table 8.3
A Continuum of the Degree of Overrepresentation in the Proposed USA Territorial-style Seat Allocation of the Upper House in Burma in
Comparison with the Upper Houses of the Entire Universe of the Eleven Longstanding Federal Democracies
Ratio of best-represented to worst-
represented federal unit
Malapportionment in upper housesa (by population)b % of seats of best-represented decilec
Nation Index Nation Ratio Nation %
Burma .634 Burma 165/1 Burma 62.5
Argentina .485 Brazil 144/1 Argentina 44.8
Brazil .403 Argentina 85/1 Brazil 41.3
United States .364 United States 68/1 United States 39.7
Switzerland .344 Switzerland 40/1 Switzerland 38.4
Canada .340 Canada 21/1 Canada 33.4
Australia .296 Australia 13/1 Australia 28.7
Spain .285 Germany 13/1 Germany 24.0
Germany .244 India 11/1 Spain 23.7
India .074 Spain 10/1 India 15.4
Austria .030 Belgium 2/1 Austria 11.9
Belgium .015 Austria 1.5/1 Belgium 10.8
Sources: The data for Burma comes from the county’s official web page: www.myanmar.com. The data for all other countries, including the malapportionment
indexes for Canada and Belgium, comes from Alfred Stepan, ‘‘Toward a New Comparative Politics of Federalism, (Multi)Nationalism, and Democracy:
Beyond Rikerian Federalism,’’ in Arguing Comparative Politics (Oxford University Press, 2001), p. 334.
Note: The malapportionment index is computed as (1/2) 兺 |si – vi|, where si = % of seats and vi = % of population. This calculation follows the formula
proposed by David Samuels and Richard Snyder in ‘‘The Value of a Vote: Malapportionment in Comparative Perspective,’’ British Journal of Political Science,
2001.
a
If there were perfect one-person, one-vote representation in the upper house, the index of malapportionment would be zero. In the worst possible case,
where the least populated state has all the seats in the upper chamber, the index would be 1. In the case of Austria, each state is represented in the upper
chamber, but the total number of representatives is allocated almost according to population. In Germany, the 1949 Basic Law stipulates that all of the sub-units
(or Länder) of the federation shall have at least three votes, but that Länder with more than 2 million inhabitants shall have four, Länder with more than 6
million inhabitants five, and Länder with more than 7 million inhabitants six. On the opposite end is Burma, where equal representation is given to all states,
some of which have very small populations and one of which has an extremely large population, thus resulting in the most malapportioned upper house in the
whole country set.
b
This ratio compares the value of a vote in the least populous sub-unit of the federation with the value of a vote in the most populous sub-unit. In Germany,
the least populous Land in 1993 was Bremen, with a population of 686,000, which was allocated three votes in the upper house. The most populous Land,
North Rhine–Westphalia, had a population of 17,679,000 and was allocated six votes. Thus, one vote in Bremen was worth 13 votes in North Rhine–
Westphalia. In the case of Burma, one vote in the Kayah State, with 246,000 inhabitants, would be worth 165 votes in ‘‘Ministerial Burma,’’ with a population of
40,647,593. In this scenario, Burma’s potential best- to worst-represented state ratio is thus 165/1, which is fifteen times greater than the average of the world’s
longstanding asymmetric federation, which is 11/1.
c
In this scenario, in Burma, five states—Kachin, Kayah, Kayin, Chin, and Mon—constitute only 10.52% of Burma’s total population. However, under one
NCUB constitutional proposal—in which ‘‘Ministerial Burma’’ is one of the total of eight ethno-federal states in the federation, each of which would receive
12.5% of the upper house seats—these five states would receive 62.5% of the seats in the upper chamber and thus would constitute an overwhelming ‘‘blocking
win-set’’ for any constitutional change and also a majority in simple majority votes.
270 crafting state-nations
a population of only 246,000, making a vote in Kayah worth 165 votes in Minis-
terial Burma. Thus, insistence on the application of the U.S. formula would
enhance incentives (already high) for the military to stay in power and to resist any
free elections.
After some discussion, the negotiators came to appreciate the obstacles to any
potential democratization the U.S.-style upper chamber formula would present in
Burma. There was also a growing interest in asymmetrical federalism. This for-
mula was seen as a potential way of constitutionally embedding some of the
specific desires of different minority nationalities for protection, such as the rights
of non-Buddhists in the state of Karen (where there are sizeable numbers of
Christians, who feared persecution) or the minority-language rights for the Shan-
speaking majority in the state of Shan.
4. A high number of electorally based veto players and their constraining im-
pact on inequality-reducing legislation. Let us now analyze the question of ‘‘veto
players’’ in the U.S. model of federalism. George Tsbellis defines a veto player
as ‘‘an individual or collective actor whose agreement is required for a policy
decision.’’16 Using spatial modeling and empirical arguments, Tsbellis makes a
convincing case that the more institutional veto players there are in a political
system—unitary or federal, parliamentary or presidential—the more difficult it is
to construct a win-set to alter the political status quo, say, by passing a bill to
extend welfare coverage to previously excluded minority groups.17
All federal democracies with a multinational dimension, as we have seen, have
created a series of special arrangements to accommodate the social diversity in
their polities. However, federal democracies in multinational societies have as
much, indeed possibly more, need to produce reasonable material standards of
welfare for all its citizens as other types of democracies. But do the special arrange-
ments, accommodations, and constitutional asymmetries of Canada, Belgium,
Spain, and India make ordinary legislation exceptionally difficult, because they
produce a higher number of ‘‘veto players’’ compared to the U.S. model? Also,
how do they compare with the U.S. model of federalism on some of the standard
indicators of poverty and inequality? Is the cost of multinational accommodation
and asymmetry weak polity-wide welfare policies?
Let us first focus on veto players that are electorally based. As we have seen, in
the U.S. federal model, the approval, in separate votes, of both the lower chamber
and the upper chamber are required for legislation. Two veto players.
Tsbellis considers the U.S. president a veto player because two separate signifi-
cant supermajorities (one in each chamber) are needed to override a presidential
veto. Three veto players.
Also, unless three-fourths of the states vote to accept a new amendment to the
Constitution, that amendment cannot pass. Thus, there are four electorally based
veto players in the U.S. system for something as basic as a constitutional change:
the states, the Senate, the House of Representatives, and the president.
How many veto players are necessary, in the admittedly complex political
arrangements crafted to accommodate the social diversity of multinational soci-
eties in Spain, Belgium, Canada, and India?
Spain, Canada, Belgium, and India are all parliamentary democracies in which
the lower house has an absolute veto on legislation. One veto player.
In all four countries, the constituent units on some issues relating to their
constitutionally embedded ‘‘asymmetrically federal’’ prerogatives, have veto pow-
ers. Two veto players.
In the judgment of most analysts, the upper house in these four countries
cannot exercise an absolute veto.18 Still only two veto players.
Should we count the prime minister as a veto player? In the technical literature
on counting veto players, the prime minister’s blocking capacity is counted as
being ‘‘absorbed’’ within the parliament’s veto power, so it does not count as a
separate veto. The (correct) reasoning behind this counting convention is that the
prime minister’s mandate can be revoked by a hostile vote of a simple majority of
the parliament. In that sense, unlike a directly elected president of the U.S. sort, a
prime minister does not have a fixed and independent mandate and cannot
sustain a long conflict with the majority in the parliament or even within his or
her party. Grand total: two veto players.
Thus, none of the admittedly complex, multinational, parliamentary, asym-
metrically federal democratic polities of Spain, Canada, India, or Belgium
have more than two ‘‘electorally based’’ veto players. In fact, when we look at
the entire universe of the world’s twenty-three longstanding democracies from
the set of advanced OECD member economies, the United States is the only
country in the set to have more than three ‘‘electorally based’’ veto players (see
table 8.4).
How does having a high number of veto players correlate with poverty and
18. See Stepan, ‘‘Electorally Generated Veto Players in Unitary and Federal Systems,’’ in Fed-
eralism and Democracy in Latin America, ed. Edward L. Gibson (Baltimore: Johns Hopkins Univer-
sity Press, 2004).
272 crafting state-nations
table 8.4
Classification of the Number of Electorally Based Institutional Veto Players in the Universe of the
Twenty-three Longstanding Advanced Economy OECD Democracies
State structure
Democracies with Federal; presidential, bicameral with both houses having extensive veto powers; mem-
four institutional ber states have absolute veto over proposals to amend the Constitution unless three-
veto players fourths of them ratify the amendment (United States)
Democracies with Federal; parliamentary, or collective executive generated by parliament, upper cham-
three institutional ber has veto; frequent referendums in which a law passed by both houses can be ve-
veto players toed unless a majority of the cantons or member states approve the law (Switzerland
and Australia)
Democracies with Federal; parlia- Asymmetrical fed- Unitary state; bi- Unitary state; par-
two institutional mentary, bi- eral; parliamen- cameral but upper liamentary, bi-
veto players cameral, with tary, bicameral chamber has weak cameral where
upper chamber with weak upper veto; semi- upper chamber
veto; member chamber veto presidential sys- has some veto ca-
states exercise veto power, but regions tem when the pacity (Italy,
power only have some consti- president does not Japan, Nether-
through upper tutionally embed- command a legis- lands)
chamber (Ger- ded veto powers lative majority
many) (Belgium, Spain, (France during
Canada) ‘‘cohabitation’’)
Democracies with Federal; parlia- Unitary state; par- Unitary state; par- Unitary state; par-
one institutional mentary, bi- liamentary, bi- liamentary, bi- liamentary, uni-
veto player cameral, where cameral, but cameral but upper cameral (Finland,
upper chamber upper chamber chamber lacks Greece, Lux-
and member has weak veto, or veto (Ireland, embourg, Portu-
states have no veto semi-presidential Sweden, Norway, gal, New Zealand,
(Austria) system when the U.K.) Denmark, Ice-
president controls land)
a majority in both
houses (France
during ‘‘non-
cohabitation’’)
Sources: Alfred Stepan, ‘‘Electorally Generated Veto Players in Unitary and Federal Systems,’’ in Federalism
and Democracy in Latin America, ed. Edward L. Gibson (Baltimore: Johns Hopkins University Press, 2004),
pp. 323–364, esp. 325–333; George Tsebelis and Jeannette Money, Bicameralism (New York: Cambridge
University Press, 1997).
19. World Development Indicators: 2007 (Washington, DC: World Bank, 2007), pp. 66–67.
274 crafting state-nations
table 8.5
Veto Players and Inequality in OECD Advanced Democracies
Fairness of
financial
Average Gini Average poverty contribution to
index rate of children of Average health system
(the higher the ‘‘single mothers’’ percentage of (the higher the
number, the after all population over number, the
greater the government 60 years old living greater the
inequality) transfers in poverty unfairness)
Sources: For Gini index figures, see World Development Indicators: 2010 (Washington, D.C.: World Bank,
2010), pp. 94–96. For poverty rate in children of ‘‘single mother’’ families, see Luxembourg Income Study, as
discussed in Lee Rainwater, ‘‘Legality and Poverty in Comparative Perspective,’’ p. 16. For population over 60
living in poverty, see Anthony Atkinson, Lee Rainwater, and Timothy Smeeding, Income Distribution in
OECD Countries (Paris: OECD, 1995), p. 104. For fairness rankings, see World Health Organization, The
World Health Report: 2000, Annex Table 7, p. 188. For the sources on the number of veto players, see table 8.4.
table 8.6
USA Inequality Indicators in Comparison to Other OECD Federal States and Entire OECD Set
USA rank against other USA rank against entire
federal countries in the set unitary and federal set
Sources: See Luxembourg Income Study, as discussed in Lee Rainwater, ‘‘Legality and Poverty in
Comparative Perspective,’’ p. 16, and the sources indicated in tables 8.4 and 8.5. Robert A. Dahl has a welfare
state index and a social expenditures index for eighteen democracies. The United States is the worst or
second-worst on each index. Dahl, How Democratic Is the American Constitution? p. 169; United Nations
Development Programme, ‘‘Human Development Report, 2009,’’ http://hdr.undp.org/eng/reports/
global/hdr2009/.
the u.s. federal model and multinational societies 275
societies, U.S.-style federalism, far from being the best possible formula, is likely to
be a highly constraining—and possibly the worst—federal governance formula for
old, new, and possible future democracies.
conclusion
Alcock, Anthony Evelyn. The History of the South Tyrol Question. Geneva: Graduate
Institute of International Studies, 1970.
Alemchiba, M. A Brief Historical Account of Nagaland. Kohima, India: Naga Institute
of Culture, 1970.
Almond, Gabriel A. ‘‘Myron Weiner on India and the Theory of Democratization.’’ In
India and the Politics of Developing Countries: Essays in Memory of Myron Weiner.
Edited by Ashutosh Varshney. New Delhi: Sage, 2004, pp. 29–42.
Anderson, Benedict. Imagined Communities: Reflections on the Origin and Spread of
Nationalism. London: Verso, 1983.
Arel, Dominique. ‘‘The Hidden Face of the Orange Revolution: Ukraine in Denial
towards Its Regional Problem.’’ Translation of ‘‘La face cachée de la Révolution
Orange: l’Ukraine en négation face á son problèm régional.’’ Revue d’études com-
paratives Est-Ouest 37 (December 2006).
————. ‘‘Parliamentary Blocs in the Ukrainian Supreme Soviet.’’ Journal of Soviet Na-
tionalities 1, no. 4 (Winter 1991), pp. 108–154.
————. ‘‘Ukraine Since the War in Georgia.’’ Survival 50, no. 6 (December 2008–
January 2009), pp. 15–25.
————. ‘‘Ukraine: The Temptation of the Nationalizing State.’’ In Political Culture and
Civil Society in Russia and the New States of Eurasia. Edited by Vladimir Tisma-
neanu. Armonk, NY: M. E. Sharpe, 1995, pp. 157–188.
Arel, Dominique, and Valeri Khmelko. ‘‘Regional Divisions in the 2004 Presidential
Elections in Ukraine.’’ Paper prepared for the Danyliw Research Seminar in Con-
temporary Ukrainian Studies at the Chair of Ukrainian Studies. University of Ot-
tawa, September 29–October 1, 2005
Arora, Balveer. ‘‘Negotiating Differences: Federal Coalitions and National Cohesion.’’
In Transforming India: Social and Political Dynamics of Democracy. Edited by
Francine R. Frankel, Zoya Hasan, Rajeev Bhargava, and Balveer Arora. Oxford and
Delhi: Oxford University Press, 2000, pp. 176–206.
————. ‘‘Spécificité Ethnique, Conscience Régionale et Développement National:
Langues et Fédéralisme en Inde.’’ Unpublished Ph.D. dissertation, Fondation Na-
tionale des Sciences, Politiques, Paris, 1972.
278 bibliography
Aspinall, Edward. ‘‘Aceh: Two Years of Peace.’’ Inside Indonesia, no. 90 (October–
December 2007).
————. ‘‘From Islamism to Nationalism in Aceh, Indonesia.’’ Nations and Nationalism
13, no. 2 (October 2007), pp. 245–263.
————. ‘‘The Helsinki Agreement: A More Promising Basis for Peace in Aceh?’’ Policy
Studies 20 (Washington, DC: East-West Center, 2005).
————. Islam and Nation: Separatist Rebellion in Aceh Indonesia. Stanford: Stanford
University Press, 2009.
————. ‘‘Modernity, History, and Ethnicity: Indonesian and Acehnese Nationalism in
Conflict.’’ Review of Indonesian and Malaysian Affairs 36 (2002), pp. 3–33.
Aspinall, Edward, and Harold Crouch. ‘‘The Aceh Peace Process: Why It Failed.’’
Policy Studies 1 (Washington, DC: East-West Center, 2003).
Bardhan, Pranab. The Political Economy of Development in India. New Delhi: Oxford
University Press, expanded edition 1998.
Barnett, Marguerite Ross. The Politics of Cultural Nationalism in South India. Prince-
ton: Princeton University Press, 1976.
Barrington, Lowell W., and Erik S. Herron. ‘‘One Ukraine or Many? Regionalism in
Ukraine and its Political Consequences.’’ Nationalities Papers 32 (March 2004),
pp. 53–86.
Barros, James. The Aland Islands Question: Its Settlement by the League of Nations.
New Haven: Yale University Press, 1968.
Barry, Brian. Culture and Equality: An Egalitarian Critique of Multiculturalism. Cam-
bridge, MA: Polity Press, 2001.
Baruah, Sanjib. ‘‘Confronting Constructionism: Ending India’s Naga War.’’ Journal of
Peace Research 40 (2003), pp. 321–338.
Basta, Lidija. ‘‘Minority and Legitimacy of a Federal State.’’ In Federalism and Multi-
ethnic States: The Case of Switzerland. Edited by Lidija Basta and Thomas Fleiner.
Fribourg, Switzerland: Institute of Federalism, 1996, pp. 41–69.
Bauböck, Rainer. ‘‘United in Misunderstanding: Asymmetry in Multinational Federa-
tions.’’ ICE Working Paper Series. Austrian Academy of Sciences 26 (May 2002).
BBC News. ‘‘French MPs Back Corsica Devolution.’’ May 22, 2001. Available at http:/ /
news.bbc.co.uk/2/hi/europe/1344285.stm.
Bechert, Heinz. ‘‘S.W.R.D. Bandaranike and the Legitimation of Power through Bud-
dhist Ideals.’’ In Religion and Legitimation of Power in Sri Lanka. Edited by Bard-
well L. Smith. Chambersburg, PA: Anima Books, 1978, pp. 199–211.
Benhabib, Seyla. The Claims of Culture: Equality and Diversity in the Global Era.
Princeton: Princeton University Press, 2002.
Beyme, Klaus von. ‘‘Die Asymmetrisierung des postmodernen Föderalismus.’’ In Die
Reformierbarkeit der Demokratie und Blockaden. Edited by Renate Mayntz and
Wolfgang Streek. Frankfurt/Main: Campus, 2003, pp. 239–258.
Bhargava, Rajeev. ‘‘The Distinctiveness of Indian Secularism.’’ In The Future of
Secularism. Edited by T. N. Srinivasan. Oxford: Oxford University Press, 2006,
pp. 20–53.
bibliography 279
Chhibber, Pradeep, and Ken Kollman. The Formation of National Party Systems:
Federalism and Party Competition in Britain, Canada, India, and the U.S. Prince-
ton: Princeton University Press, 2004.
Colley, Linda. Britons: Forging the Nation, 1707–1837. New Haven: Yale University
Press, 1992.
Colton, Timothy J., and Cindy Skach. ‘‘The Predicament of Semi-Presidentialism.’’ In
Democracies in Danger. Edited by Alfred Stepan. Baltimore: Johns Hopkins Uni-
versity Press, 2009, pp. 121–136.
————. ‘‘The Russian Predicament.’’ Journal of Democracy 16 (July 2005), pp. 113–126.
Constituent Assembly Debates. Volume 2. New Delhi, India, 1951.
Cotta, Maurizio. ‘‘Elite Unification and Democratic Consolidation in Italy: An His-
torical Overview.’’ In Elites and Democratic Consolidation in Latin America and
Southern Europe. Edited by John Higley and Richard Gunther. Cambridge: Cam-
bridge University Press, 1992, pp. 146–177.
Cowell, Alan. ‘‘Greenland Vote Favors Independence.’’ The New York Times, Novem-
ber 26, 2008. Available at www.nytimes.com/2008/11/27/world/europe/27green
land.html, accessed November 30, 2008.
Crowther, William. ‘‘The Politics of Ethno-National Mobilization: Nationalism and
Reform in Soviet Moldavia.’’ The Russian Review 50 (April 1991), pp. 183–202
Cuaz, Marco. ‘‘La Valle d’Aosta. Un’identità di frontiera fra Italia, Europa ed etno-
nazionalismi.’’ In Altre Italie: Identità nazionali e Regioni a statuto speciale. Edited
by Gaspare Nevola. Rome: Carocci editore, 2003, pp. 1–18.
Cunningham, David. ‘‘Veto Players and Civil War Duration.’’ American Journal of
Political Science 50, no. 4 (October 2006), pp. 875–892.
D’Anieri, Paul. ‘‘Ethnic Tensions and State Strategies: Understanding the Survival
of the Ukrainian State.’’ Journal of Communist Studies and Transition Politics 23
(March 2007), pp. 4–29.
Daftary, Farimah. ‘‘Insular Autonomy: A Framework for Conflict Resolution? A Com-
parative Study of Corsica and the Alland Islands.’’ Global Review of Ethnopolitics 1,
no. 1 (September 2001), pp. 19–40.
Dahl, Robert. Democracy, Liberty, and Equality. Oslo: Norwegian University Press,
1986.
————. How Democratic Is the American Constitution? New Haven: Yale University
Press, 2001.
Dang, Satyapal. Genesis of Terrorism: Analytical Study of Punjab Terrorists. New Delhi:
Patriot Publishers, 1988.
Das Gupta, Jyotirindra. Language Conflict and National Development: Group Poli-
tics and National Language Policy in India. Berkeley: University of California
Press, 1970.
Debes, Hans Jacob. ‘‘The Formation of a Nation: The Faroe Islands.’’ In Ethnicity and
Nation Building in the Nordic World. Edited by Sven Tägil. London: Hurst, 1995,
pp. 63–84.
Denmark. Act no. 473 of 12 June 2009, Act on Greenland Self-Government.
bibliography 281
Desai, A. R., ed. Violation of Democratic Rights in India. Bombay: Popular Prakashan,
1986.
Diamond, Larry. ‘‘Economic Development and Democracy Reconsidered.’’ In Re-
Examining Democracy. Edited by Gary Marks and Larry Diamond. Newbury Park,
CA: Sage, 1992, pp. 93–139.
Diamond, Larry, Juan J. Linz, and Seymour Martin Lipset, ed. Democracy in Develop-
ing Countries. Boulder, CO: Lynne Rienner, 1989.
Di Michele, Andrea, Francisco Palermo, and Günther Pallaver, eds. 1992: Fine di Un
Conflitto: Dieci anni dalla chiusura della questione sudtirolese. Bologna: il Mulino,
2003.
Dirks, Nicholas B. Castes of Mind: Colonialism and the Making of Modern India.
Princeton: Princeton University Press, 2001.
Dominici, Thierry. ‘‘Le système partisan nationalitaire corse contemporain: étude
d’un phénomène politique.’’ BCN Political Science Debates 4 (2005), pp. 43–88.
The Economist. ‘‘Chronicle of a Death Foretold.’’ December 1, 2009.
————. ‘‘A War as Strange as Fiction.’’ June 7, 2007.
Elazar, Daniel. Federal Systems of the World: A Handbook of Federal, Confederal, and
Autonomy Arrangements. Essex, UK: Longman, 1991.
Eldersveld, Samuel, and Bashiruddin Ahmed. Citizens and Politics: Mass Political
Behavior in India. Chicago: University of Chicago Press, 1978.
Embree, Ainslie. India’s Search for National Identity. Delhi: Chanakya Publications,
1980.
Euskalherria en la encuesta Europea de valores. Bilbao: Universidad de Deusto, 1992
Faris, Stephan. ‘‘Ice Free: Will Global Warming Give Greenland Its Independence?’’
New York Times Magazine, July 27, 2008.
‘‘Faroe Islanders Head to the Polls Saturday.’’ Agence France-Presse, January 17, 2008.
Fenet, Alain. ‘‘Difference Rights and Language in France.’’ In Language, Nation, and
State: Identity Politics in a Multilingual Age. Edited by Tony Judt and Dennis
Lacorne. New York: Palgrave/Macmillan, 2004, pp. 19–62.
‘‘Finland: Aaland Islands.’’ In Federal Systems of the World. Edited by Daniel J. Elazar.
Essex: Longman, 1991.
Flora, Peter, ed. State, Economy, and Society in Western Europe, 1815–1975: A Data
Handbook in Two Volumes. Frankfurt: Campus Verlag, 1983.
Foighel, Isi. ‘‘Home Rule in Greenland 1979.’’ Nordic Journal of International Law 48
(1979).
Forrester, Duncan B. ‘‘Kamaraj: A Study in Percolation of Style.’’ Modern Asian Stud-
ies 4, no. 1 (1970), pp. 43–61.
————. ‘‘The Madras Anti-Hindi Agitation, 1965: Political Protest and its Effect on
Language Policy in India.’’ Pacific Affairs 39 (Spring–Summer 1966), pp. 19–36.
France. Constitutional Council. Decision 2001–454 DC of 17 January 2002.
‘‘The Freedom House Survey for 2009.’’ Journal of Democracy 19, no. 2 (April 2010).
Freedom House. Freedom in the World: Political Rights and Civil Liberties. New York:
Freedom House, multiple years.
282 bibliography
Friend, Robert C. The Italian Prefects: A Study in Administrative Politics. New Haven:
Yale University Press, 1963.
Frye, Timothy. ‘‘A Politics of Institutional Choice: Post Communist Presidencies.’’
Comparative Political Studies 30, no. 4 (1997), pp. 523–552.
Furet, François, and Mora Ozouf, eds. A Critical Dictionary of the French Revolution.
Cambridge: Belknap Press, 1989.
Gandhi, M. K. Collected Works. New Delhi: Publications Division, Government of
India, 1969.
General Election Study, Belgium 1995. Interuniversitair Steunpunt Politieke-Opinie-
onderzoek, K. U. Leuven and Point d’Appui Interuniversitaire sur l’Opinion Pub-
lique et la Politique, U. C. Louvain, 1998.
Gellner, Ernest. Nations and Nationalism. Oxford: Oxford University Press, 1983.
Gill, K. P. S. Punjab, The Knights of Falsehood. New Delhi: Har-Anand Publications,
1997.
Gombrich, Richard, and Gananath Obeyesekere. Buddhism Transformed: Religious
Change in Sri Lanka. Princeton: Princeton University Press, 1988.
Goswami, B. B. The Mizo Unrest: A Study of Politicization of Culture. Jaipur: Aalekh
Publishers, 1979.
Gunasinnghe, Newton. ‘‘Community Identity and Militarization in Sri Lanka: Sri
Lankan Armed Forces.’’ In The Challenge in South Asia: Democracy, Development,
and Regional Cooperation. Edited by Ponna Wijnaraja and Akmal Hussain. New
Delhi: Sage, 1989. 243–249.
Habermas, Jürgen. Between Facts and Norms. Cambridge, MA: MIT Press, 1998.
————. Die Einbeziehung des Anderen: Studien zur politschen Theorie. Frankfurt am
Main: Suhrkamp, 1996.
Habib, Irfan. ‘‘Emergence of Nationalities.’’ Social Scientist, no. 37 (August 1975),
pp. 14–20.
Hale, Henry E. Foundations of Ethnic Politics. Cambridge: Cambridge University
Press, 2008.
Hardgrave, Richard L. Jr. ‘‘Religion, Politics, and the DMK.’’ In South Asian Politics
and Religion. Edited by Donald Eugene Smith. Princeton: Princeton University
Press, 1966.
————. ‘‘The Riots in Tamilnad: Problems and Prospects of India’s Language Crisis.’’
Asia Survey 5 (August 1965), pp. 399–407.
Harhoff, Frederick. ‘‘Greenland’s Withdrawal from the European Communities.’’
Common Market Law Review 20 (1983), pp. 13–33.
Hasan, Mushirul. Legacy of a Divided Nation: India’s Muslims since Independence.
Delhi: Oxford University Press, 1997.
Hazarika, Sanjoy. Strangers of the Mist: Tales of War and Peace from India’s Northeast.
London: Penguin, 1994.
The Hindu. ‘‘How India Voted: Verdict 2009.’’ March 26, 2009.
Hirschman, Albert. Exit, Voice, and Loyalty: Responses to Declines in Firms, Organiza-
tions, and States. Cambridge: Harvard University Press, 1972.
Hooghe, Liesbet. ‘‘Belgium: Hollowing the Center.’’ In Federalism and Territorial
bibliography 283
Cleavages. Edited by Ugo M. Amoretti and Nancy Bermeo. Baltimore: Johns Hop-
kins University Press, 2004, pp. 55–92.
Horowitz, Donald L. ‘‘Incentives and Behavior in the Ethnic Politics of Sri Lanka and
Malaysia.’’ Third World Quarterly 11 (October 1989), pp. 18–35.
Hrytsak, Yaroslav. ‘‘National Identities in Post-Soviet Ukraine: The Case of Lviv and
Donetsk.’’ In Cultures and Nations of Central and Eastern Europe. Edited by Zvi
Gitelman, Lubomyr Hajda, John-Paul Himka, and Roman Solchanyk. Harvard
Ukrainian Studies 22 (1998), pp. 263–282.
————. ‘‘On the Relevance and Irrelevance of Nationalism in Ukraine.’’ Second An-
nual Cambridge-Stasiuk Lecture on Ukraine, February 20, 2004.
————. ‘‘Ukrainian Nationalism, 1991–2001: Myths and Perceptions.’’ Austrian Institute
for Eastern and South-Eastern Europe. Vienna, Austria, October 15, 2001.
Human Rights Watch. ‘‘War on the Displaced: Sri Lankan Army and LTTE Abuses
against Civilians in the Vanni.’’ February 2009.
‘‘Interactive Ukraine Election Maps.’’ Kyiv Post. January 18, 2010. Available at www
.kyivpost.com/news/politics/detail/57376/.
International Crisis Group. ‘‘Aceh: Post Conflict Complications.’’ Asia Report, no. 139,
October 4, 2007.
————. ‘‘Aceh: So Far, So Good.’’ Asia Briefing, no. 44, December 13, 2005.
————. ‘‘Indonesia: Pre-Election Anxieties in Aceh.’’ Asia Briefing, no. 81, September 9,
2008.
International Monetary Fund. Government Finance Statistics. Washington, DC: Inter-
national Monetary Fund, various issues.
Irschick, Eugene F. Politics and Social Conflict in South India: The Non-Brahmin
Movement and Tamil Separatism, 1916–1929. Berkeley: University of California
Press, 1969.
Jaffrelot, Christophe. ed. Hindu Nationalism: A Reader. Princeton: Princeton Univer-
sity Press, 2007.
————. The Hindu Nationalist Movement in India. London: Hurst, 1996.
Jennings, Ivor. The Constitution of Ceylon. 3rd ed. London: Oxford University Press,
1953.
Jodhka, Surinder S. ‘‘Looking Back at the Khalistan Movement: Some Recent Re-
searches on its Rise and Decline.’’ Economic and Political Weekly, vol. 36, no. 16,
April 21–27, 2001, pp. 1311–1318.
Johannson, Andreas. ‘‘The Transnistrian Conflict after the 2005 Moldovan Parliamen-
tary Elections.’’ Journal of Communist Studies and Transition Politics 22 (December
2006).
Jones, Matthew. ‘‘Faroese Explore Prospects for New Economic Base.’’ Financial
Times, July 10, 2001.
Jussila, Osmo. From Grand Duchy to Modern State: A Political History of Finland since
1809. London: Hurst, 1999.
Kailash, K. K. ‘‘Coalitions in a Parliamentary Federal System: Parties and Govern-
ments in India.’’ Unpublished Ph.D. dissertation, Jawaharlal Nehru University,
2003.
284 bibliography
Kalinina, Svitlana, Alexander Lyakh, Galina Savchenko, and Adam Swain. ‘‘Regional
‘Lock-in’ or Local Hegemonic Bloc? Industrial Restructuring in the Ukrainian
Donbas.’’ Paper prepared for the Danyliw Research Seminar in Contemporary
Ukrainian Studies at the Chair of Ukrainian Studies. University of Ottawa, Septem-
ber 29–October 1, 2005.
Karat, Prakash. ‘‘Theoretical Aspects of the National Question.’’ Social Scientist 37
(August 1975), pp. 5–13.
Kashmir: An Enquiry into the Healing Touch. Report by the Association for Demo-
cratic Rights (ADR), Punjab, Human Rights Forum (HRF), Andhra Pradesh and
Organisation for Protection of Democratic Rights (OPDR), Hyderabad, India,
2003.
Kaviraj, Sudipta. ‘‘The Imaginary Institution of India.’’ Subaltern Studies 7. Edited by
Partha Chatterjee and Gyananendra Pandey. Delhi: Oxford University Press, 1993,
pp. 1–39.
————., ed. Politics in India. Oxford and Delhi: Oxford University Press, 1997.
————. The Unhappy Consciousness: Bankimchandra Chattopadhyay and the Forma-
tion of Nationalist Discourse in India. Delhi: Oxford University Press, 1995.
Keane, John. The Life and Death of Democracy. London: Simon and Schuster, 2009.
Kearney, Robert N. Communalism and Language in the Politics of Ceylon. Durham,
NC: Duke University Press, 1967.
Khan, Rasheeduddin, ed. Rethinking Indian Federalism. Shimla: Indian Institute of
Advanced Study, 1997.
Khilnani, Sunil. The Idea of India. London: Hamish Hamilton, 1997.
Kingsbury, Damien. Peace in Aceh: A Personal Account of the Helsinki Peace Process.
Jakarta: Equinox Publishing, 2006.
Kogan, Norman. A Political History of Italy: The Post-War Years. New York: Praeger,
1983.
Kohli, Atul, ed. The Success of India’s Democracy. Cambridge: Cambridge University
Press, 2001.
Kolstø, Pȧl, and Andrei Malgin. ‘‘The Transnistrian Republic: A Case of Politicized
Regionalism.’’ Nationalities Papers 26, no.1 (1998).
Kothari, Rajni. Politics in India. Boston: Little Brown, 1970.
————, ed. State and Nation Building : A Third World Perspective. Bombay: Allied, 1976
Kumar, Ram Narayan, Amrik Singh, Ashok Agrwaal, and Jaskaran Kaur. Reduced to
Ashes: The Insurgency and Human Rights in Punjab. New Delhi: South Asia Forum
for Human Rights, 2003.
Kumar, Ravinder. ‘‘India: A ‘Nation-State’ or ‘Civilisation-state’?’’ Journal of South
Asian Studies 25, no. 2, pp. 13–32.
Kumar, Sanjay. ‘‘Religious Practices among Hindus: Does This Influence Their Politi-
cal Choices?’’ CSDS Working Paper, Delhi, June 2008
Kuru, Ahmet. Secularism and State Policies toward Religions: The United States,
France, and Turkey. Cambridge: Cambridge University Press, 2009.
————. ‘‘Secularism, State Policies, and Muslims in Europe: Analyzing French Excep-
tionalism.’’ Comparative Politics 4 (2008), pp. 1–19.
bibliography 285
Luithi, Luingam, and Nandita Haksar. Nagaland File: A Question of Human Rights.
New Delhi: Lancer, 1986.
Lutz, Donald S. ‘‘Toward a Theory of Constitutional Amendment.’’ American Political
Science Review 88, no. 2 (June 1994), pp. 355–370.
Luverà, Bruno. ’’L’Euregio tirolese. Tra regionalismo transfrontaliero e micronazio-
nalismo di confine.’’ In Altre Italie. Edited by G. Nevola. Rome: Carocci, 2003,
pp. 19–33.
Lyngdoh, James. The Chronicle of an Impossible Election: The Election Commission
and the 2002 Jammu and Kashmir Assembly Elections. Delhi: Viking India, 2004.
Madeley, John T. S. ‘‘A Framework for the Comparative Analysis of Church-State
Relations in Europe.’’ West European Politics 26, no. 1 (January 2003), pp. 23–50.
Manor, James. The Expedient Utopian: Bandaranike and Ceylon. Cambridge: Cam-
bridge University Press, 1989.
Marshall, Geoffery. Constitutional Conventions: The Rules and Forms of Political Ac-
countability. Oxford: Oxford University Press, 1984.
Marshall, Monty G. ‘‘Major Episodes of Political Violence: 1946–2006.’’ Center for
Systemic Peace, 2009.
Marwah, Ved. Uncivil Wars: Pathology of Terrorism in India. New Delhi: Harper
Collins, 1995.
Masani, R. P. Dadabhai Naoroji: The Grand Old Man of India. London: Unwin
Brothers, 1939.
Maxwell, Kenneth. ‘‘Regime Overthrow and the Prospects for Democratic Transi-
tion.’’ In Transitions from Authoritarian Rule: Southern Europe. Edited by Guil-
lermo O’ Donnell, Philippe C. Schmitter, and Laurence Whitehead. Baltimore:
Johns Hopkins University Press, 1986, pp. 109–137.
————. ‘‘The Thorns of the Portuguese Revolution.’’ Foreign Affairs 54 (January 1976),
pp. 250–270.
McCall, A. G. Lushei Chrysalis. London: Luzac, 1949.
McRae, Kenneth D. Conflict and Compromise in Multilingual Societies, vol. 1. On-
tario, Canada: Wilfred Laurier University Press, 1986.
Meyer, David John. ‘‘Ethnic Territorial Autonomy and Post-Soviet Ethnic Political
Mobilization.’’ Unpublished Ph.D. dissertation, Department of Political Science,
Columbia University, 2007.
Miley, Thomas Jeffrey. Nacionalismo y política lingüística: el caso de Cataluña.
Madrid: CEPC, 2006.
Mill, John Stuart. Considerations on Representative Government (1861). In Utilitarian-
ism, On Liberty, Considerations on Representative Government. Edited by Geraint
Williams. London: Everyman, 1993.
Mitta, Manoj, and H. S. Phoolka, When a Tree Shook Delhi: The 1984 Carnage and Its
Aftermath. New Delhi: Roli Books, 2007.
Modeen, Tore. ‘‘Åland Islands.’’ In Encylopedia of Public International Law, vol. 12.
Edited by Rudolf Berhardt. Amsterdam: North Holland, 1990.
————. ‘‘The International Protection of the National Identity of the Åland Islands.’’
Scandinavian Studies in Law (1973), pp. 177–210.
288 bibliography
————. ‘‘Visiting the Vavuniya IDP Camps: An Uplifting Experience.’’ Interview with
N. Ram. The Hindu, July 4, 2009.
Rangasami, Amritha. ‘‘Mizoram: Tragedy of Our Own Making.’’ Economic and Politi-
cal Weekly, vol. 13, no. 15, April 15, 1978, pp. 653–662.
Rao, R. V. Ramachandrasekhara, ed. Indian Unity: A Symposium. New Delhi: Pub-
lication Division, Government of India, 1969.
Raz, Joseph. Ethics in the Public Domain: Essays in the Morality of Law and Politics.
Oxford: Clarendon Press, 1994.
————. The Morality of Freedom. Oxford: Oxford University Press, 1986.
Reid, Anthony, ed. Veranda of Violence: The Background to the Aceh Problem. Singa-
pore: Singapore University Press, 2006.
Renner, Michael. ‘‘The Casualities of Sri Lanka’s Intensifying War.’’ World Watch
Institute. July 18, 2007.
Requejo Coll, Ferran. Federalisme, per a què? Valencia: Tres i Quatre, 1998.
Reynolds, Andrew, Alfred Stepan, Zaw Oo, and Stephen Levine. ‘‘How Burma Could
Democratize.’’ Journal of Democracy 12, no. 4 (October 2001), pp. 94–108.
Rezvani, David A. ‘‘Federacy: The Dynamics of Semi-Sovereign Territories.’’ Un-
published D. Phil in Politics, University of Oxford, 2004.
Riker, William. The Development of American Federalism. Boston: Kluwer Academic
Publishers, 1987.
————. ‘‘Federalism.’’ In Handbook of Political Science, vol 5. Edited by Fred Green-
stein and Nelson W. Polsby. Reading, MA: Addison-Wesley, 1975, pp. 93–72.
————. Federalism: Origin, Operation, Significance. Boston: Little Brown, 1964.
Roper, Steven. ‘‘Are All Semipresidential Regimes the Same? A Comparison of
Premier-Presidential Regimes.’’ Comparative Politics 34 (April 2002), pp. 253–272.
Rose, Richard. ‘‘Divisions within Ukraine: A Post-Election Opinion Survey.’’ Studies
in Public Policy 403 (2005).
————. ‘‘Russians in the Baltic: A 1991 Survey.’’ Studies in Public Policy 287 (1991).
————. ‘‘Russians outside Russia.’’ Studies in Public Policy 283 (1991).
Rose, Richard, and Christian Haerpfer. ‘‘New Democracies Barometer V: A Twelve
Nation Survey.’’ Studies in Public Policy, no. 306 (1998).
Roux, Christophe. ‘‘The Partito Sardo D’Azione: Regionalist Mobilization in South-
ern Italy.’’ In Autonomist Parties in Europe: Identity Politics and the Revival of
Territorial Cleavage, vol. 2. Edited by Lieven De Winter, Margarita Gómez-Reino,
and Peter Lynch. Barcelona: Institut de Ciències Polítiques i Socials, 2006.
Rubin, Barnett R., and Jack Snyder, eds. Post-Soviet Political Order: Conflict and State
Building. New York: Routledge, 1998.
Rudolph, Lloyd I. ‘‘Urban Life and Populist Radicalism: Dravidian Politics in Madras.’’
Journal of Asian Studies 20 (May 1961).
‘‘Russia to Sign Deal on Military Base in South Ossetia on Wednesday.’’ Ria Novosti,
April 6, 2010.
Rustow, Dankwart. ‘‘Transitions to Democracy: Toward a Dynamic Model.’’ Compara-
tive Politics 2 (April 1970), pp. 337–363.
Samarasinghe, S. W. R. de A. ‘‘Ethnic Representation in Central Government Em-
bibliography 291
The letter f following a page number denotes a figure, and the letter t denotes a table.
Abdullah, Sheikh (of Kashmir), 112–13 25–26, 260; democracy in, 36–38,
Aceh: federacy in, 6, 229, 243–52, 255; 46n16, 264f, 272t; inequality of represen-
GAM in, 242–52; LOGA in, 251, tation in, 260t, 268t; national pride in,
252nn123–24; and MoU treaty, 247–49, 58t; as nation-state, viii, 3, 36–38, 38t;
251, 252n123; Muslims in, 242; and peace symmetrical federalism in, 260–61,
with Indonesia, xi, 242–53; shari’a law 264f; trust in institutions of, 38t; U.S.
in, 250; terrorist camps in, 251; tsunami federalism and, 258; veto in, 260, 272t
in, 244, 249; vs. the UN, 248 Austria: democracy in, 36–38, 46n16, 264f,
Advani, L. K., 83 272t; independence of, 27; inequality of
Ahtisarri, Martti, 246–48, 252n123 representation in, 260t, 268–69t; as irre-
Akali Dal party. See under Sikhs dentist, 229; national pride in, 58t; as
Åland Islands. See Finland; Sweden nation-state, viii, 3, 11, 36–38, 38t; South
Ambedkar, B. R., 69, 120–21, 138, 139t Tyrol and, 229–36, 253; symmetrical
Anderson, Benedict, 116–17 federalism in, 260–61, 264f; trust in in-
Annadurai, C. N., 118–19, 123, 129–30 stitutions of, 38t, 74f, 76t; U.S. federal-
Arel, Dominique, 179–80, 183, 183n32, 191 ism and, 258–61; veto and, 121, 259, 272t;
Argentina, viii, 36–38, 38t, 46n16, 58, 58t, Yugoslavia and, 253
74, 76t, 264f, 267, 268t authoritarianism, 64–66, 65t, 66t, 70t, 82,
Arunachalam, Ponnambalam, 148 92, 174n1, 180, 196t, 202, 228–31
Aspinall, Edward, 245 Azores. See Portugal
asymmetrical federalism: collective rights
and, 18; cultural nationalism and, 21; Badal, Prakash Singh, 92–94, 98
dangers of, 172–73; defined, 18; in de- Baltic states, 25n28, 182, 193n63. See also
mocracies, 5; in holding together Estonia; Latvia; Lithuania
federations, 26, 165; Muslim accom- Bandaranaike, S. W. R. D., 150–52, 155–
modation and, 61, 111; in nested policy 58, 158n40
grammar, 17–19, 22, 56, 170; robust mul- Barnett, Ross, 129–30, 132, 134
tinationalism and, 262–63; state-nations Baruah, Sanjib, 106, 108
and, 5–7, 8t, 17–18, 22, 173, 201; vs. uni- Basque Country, 2, 27, 30–32, 62, 122–
tary states, 17–18, 56, 170t, 203, 232; vs. 23n24, 254, 263
U.S. federalism, 260, 270; veto and, 272t Belgium: asymmetrical federalism in, 5–
Atkinson, Anthony B., 273 6, 9, 37, 165, 261–63, 264f, 271; democ-
Australia: amendment process in, 261– racy in, 36–38, 46n16, 125, 196t, 264f,
62n8; as coming together federation, 272t; identities in, 14, 31–36, 60–61;
298 index
cultural vs. secessionist nationalism, 18, European Union (EU), 11, 57, 193n63,
21–22, 22n21, 117–19, 125n29, 126–27. See 198–99, 217, 249–50, 254
also Tamils in India; Tamils in Sri
Lanka Faroe Islands. See Denmark–Faroe Island
Czech Republic, 74f, 198n73, 199 federacy
federacies: vs. associated states, 208–9,
Dahl, Robert, 5n4, 16, 259, 274t 253; vs. asymmetrical federalism, 208–
Dalai Lama, 202, 255 10, 252–55; characteristics of, 204–9,
Dalits. See India: marginal groups in 212–17, 221–26, 247–49; vs. confedera-
D’Anieri, Paul, 191 tions, 208; democracy and, 202; ideal
Dasgupta, Jyotirindra, 108 types of, 201–2, 204–7; identities in,
De Gasperi-Gruber Accord, 233–35, 253 252–55; irredentism and, 204–9, 252–56;
democracy: asymmetrical federalism in, 5; overview of, 201–55; presidentialism
defeated by political struggle, 244; and, 208; robust multinationalism and,
federacies and, 202; ideal types of, 8t; Is- xi; state-nations and, xi, 8t; unitary states
lam and, 41; multinational, 47, 116, 119, and, 201, 207, 254–55. See also Aceh;
122, 125, 129, 264f; vs. nation-state build- China; Corsica; Denmark; Finland;
ing, viii–x, 3, 108, 275; vs. oppressive na- France; Italy
tional policies, 23; religiosity and, 66– federalism: demos-enabling, 120–23; U.S.
72, 167, 168f; requirements for, 175–79; model of, xi–xii, 257–75. See also asym-
SDSA study of, 57n35, 61, 75, 161–64, metrical federalism; coming together
167; socioeconomic development and, federations; ethno-federalism;
41–42; state-nations and, viii, 10f, 14, 16– federacies; holding together federations;
17, 81, 108–11, 159, 193; vs. U.S. federal- symmetrical federalism
ism, 275; U.S. federal model of, xi–xii, federalism defined, 5n4, 16
257–75; veto in, 272t, 274t. See also In- Finland, 74f, 76t, 190n51, 272t
dian Hindus; Indian Muslims; robust Finland–Åland Island federacy, xi, 6, 202,
multinationalism; Sikhs; Sri Lankan/ 210–18, 246–48, 252n123, 253–55
Ceylonese Buddhists Firth, Peter, 250
demos-enabling federalism, 120–23 Flanders, 2, 32–33, 33t, 35–36, 135, 137, 263
Denmark, 68, 74f, 190n51, 206n2, 219, 258, Forrester, D. B., 128n37, 133
272t France: Åland Islands and, 212, 216; asym-
Denmark–Faroe Island federacy, xi, 6, metrical federalism in, 240; Corsica as
210, 218–25, 220n30, 222n35, 253–54 possible federacy in, 202, 228–32, 239–41;
Denmark-Greenland federacy, xi, 5–6, dual executive system in, 197–98; Italy
202, 210, 225–28, 247, 253–54 and, 229–31, 236; Muslims in, 29; as na-
Djuli, Nur, 246–47, 246n109, 252n123 tion builder, 22–23, 25; as nation-state, 2–
DMK (Dravida Munnetra Kazhagam). 3, 9, 28, 29, 52, 83, 179, 202, 239; political
See Tamil Nadu: Dravidians in parties in, 81; religion and, 19, 68–69; trust
Dravidians. See Tamil Nadu, Dravidians in in institutions of, 74f; veto in, 272t
Franco, Francisco, 23, 30, 62
Elazar, Daniel J., 209, 226, 236 Freedom House Surveys, 169, 174, 193
Elster, John, 261
England. See Great Britain; United King- GAM (Gerakan Aceh Merdeka), 242–52
dom Gandhi, Indira, 42, 74n52, 84, 92–94, 99,
Estonia, 11, 181, 181n25, 198n73, 199, 216 113, 131–33
ethno-federalism, 10, 15–17, 15n13, 19–21, Gandhi, Mahatma, ix, 52–53, 68, 123, 138–
21n19, 35, 37, 144, 144n1, 255, 269 39, 139t
300 index
82, 167, 168f; Indian institutions trusted Jammu and Kashmir, 39n1, 49n22, 90t,
by, 72t, 167t; majoritarianism of, 83, 162– 109–15, 209. See also Kashmir
63, 163t; Muslims and, 68–69, 82–83, Jammu region, 109n37, 111–12. See also
156n36; national identity among, 61, 61t; Kashmir
nationalists among, 51–52, 67, 69, 83, 85, Janata Party. See BJP (Bharatiya Janata
112, 156n36; populations of, ix, 41, 46–47, Party)
67; state support for, 156, 156n36; veto Japan, 2–3, 9, 74f, 146, 272t
power of, 54; voter turnout among, 80, 80t Jennings, Ivor, 149
Indian Muslims: accommodation of, 61; Jones, Sidney, 250, 251n119
cultural assimilation and, 24; democ-
racy supported by, ix, 69–70, 70t, 71f, Kalla, Jusuf, 245, 245n104, 252
72t, 80, 82; governmental protection of, Kashmir: elections in, 77n55, 113, 115n43;
54; Hindus and, 68–69, 82–83, 156n36; Hindus in, 109n37; India and, ix, 48, 61,
as marginal group, ix; massacres of, 109–15, 114t; languages in, 112; location
74n52, 80, 85; national identity among, of, 90t; Muslims in, 39n1, 61, 109n37,
61, 61t; national pride among, 59, 59t; 111, 113; Nehru and, 109–10; Pakistan
Partition and, 54; population of, 41, and, 39n1, 110, 113–14, 114t, 115n43; polit-
46n17, 67; poverty level of, 45t; on ical history of, 109–13; populations in,
Provincial Council Power, 166t; state in- 49n22, 109n37; regional divisions of,
stitutions supported by, 87; trust in In- 39n1, 109n37; robust multinationalism
dian Army among, 167t; veto power of, and, 2, 39n1; separatist insurgency in, x,
54; voting efficacy and, 78, 79t; voting 39, 39n1, 48, 89–90, 109, 113–14; and
rates among, 72t, 80, 80t trust in Indian institutions, 113–14; UN
Indian National Election Studies, ix, 63, and, 90, 110. See also Jammu and Kash-
70, 78, 108n35, 115n43, 134 mir; Jammu region; Ladakh region
Indian Peripheral States, 90t Kaviraj, Sudipta, 51
individual rights and collective recogni- Khalistan independence movement. See
tion, 17–19, 56 Punjab; Sikhs
Indonesia: Christians in, 242; democracy Khilnani, Sunil, 56, 56n33
and, 242–52; federacy and, 6, 229, 243– Khmelnytskyi, Bohdan, 186
52; geography of, 49; and Helsinki nego- Khrushchev, Nikita, 175
tiations, 245–53; military rule in, 75n54, Kogan, Norman, 231
242–43; MoU treaty in, 246n109, 247– Korea, vii n1, 63–64, 65n43, 65t, 75n54
49, 251, 252n123; Muslims in, 242–43; Kravchuk, Leonid, 177–78, 182, 194
Norwegian peacekeeping mission in, Krawchenko, Bohdan, 182
244–45; and peace with Aceh, xi, 242– Kuchma, Leonid, 174n1, 186, 194, 198n73,
53, 250n116; vs. Tamil Tigers, 244; terror- 199
ism in, 242; U.S. federal model and, 258
inequality of representation, 259, 260t, Ladakh region, 39n1, 109n37, 111–12, 114t.
270–75 See also Jammu region; Kashmir
Inglehart, Ronald, 37n46, 57 Laldenga (Mizo leader), 102–5
Ireland, 5n5, 42, 51n23, 74f, 263n11, 272t Lapidoth, Ruth, 210, 215, 217–18, 220n30,
irredentism. See Austria; Denmark–Faroe 221–23
Island federacy; federacies; Italy; Swe- Lasala-Blanco, Narayani, 140n55
den; Ukraine Latvia, 11, 181, 181n25, 198n73, 199, 216
Islam. See Muslims Law on Governing Aceh (LOGA), 251,
Italy, xi, 74f, 202, 216, 229–37, 253–54, 272t 252nn123–24
302 index
3, 7–9, 8t, 13, 17, 27–28, 37–38, 184; as 224, 271, 272t. See also Belgium; Can-
modern state model, 275; vs. national- ada; India; presidential systems; semi-
ism, xi, 11, 13, 23; vs. state-nations, vii– presidential systems; Spain
viii, 1–38, 88, 157–59, 173–74, 184–86, Partition, 41, 51, 54, 68, 92, 109, 109n37, 120
188–90, 243; superiority claimed for, 38; Patel, Sardar, 123, 139t
unitary, x, 3, 26, 207, 210, 212, 237, 255. Periar (Naicker, Ramaswami), 118–19,
See also Argentina; Australia; Austria; 140n53
Belgium; Brazil; Ceylon; Denmark; Phadnis, Urmila, 155
Finland; France; Germany; India; Lux- Poland, 176–77, 199–200, 216
embourg; Spain; Sri Lanka; Switzer- political integration vs. cultural assimila-
land; Ukraine tion, 18, 21
NDA (National Democratic Alliance), polity-wide careers, 20–21, 129–30, 150
83–84, 86–87 polity-wide parties, 18, 20–22, 81, 92, 123–
Nehru, Jaraharwal: death of, 130; inclu- 26, 135, 135t, 153
sionary policies of, ix; Jammu and Kash- Ponnambalam, G. G., 150
mir and, 109–10, 112; Kamaraj and, 124; Portugal, 199n74; amendment process in,
Kashmir and, 109–10, 110n38; multina- 261–62n8, 272t; Azore secession from,
tional tensions managed by, 4; on Naga- xi, 228, 237–39; Ceylon and, 153; de-
land, 101n15; name recognition of, 138, mocracy in, 202; federacy in, 202, 254; as
139t; vs. nationalism, 53; vs. Pakistan, nation-state, 3; semi-presidentialism in,
109–10; as positive symbol, 185n39; Pun- 199–200; trust in institutions of, 74f;
jab and, 92 veto in, 272t
Nehru (Motilal) Report (1928), 54–55 poverty and veto players, 271–75, 274t
nested policy grammar, viii, x, 17–22, 56, presidential systems, 264–65, 272t. See also
119–23, 167–69, 170t, 197 parliamentary systems; semi-
Netherlands, 74f, 272t presidential systems
New Zealand, 81, 261–62n8, 272t Puerto Rico, 209–10
‘‘no exit’’ policies, 48, 96, 127, 127n33 Punjab: Congress Party in, 96, 98–100;
Norris, Pippa, 73 economy of, 91n1; Hindus in, 92; hu-
Norway, 11, 73, 74f, 164, 169, 218n24, 244, man rights issues in, 95n9; identities in,
272t 51, 62; Khalistan independence move-
Nurdin, A. R., 250n116 ment in, x, 2, 39, 48–49, 89, 91–100, 111;
location of, 90f; vs. Nagaland, 96;
Obeyesekere, Gananath, 145, 154 Nehru and, 92; population of, 49n22;
Olcott, Henry Steel, 154–55 terrorist violence in, 95, 96f; and trust in
Orange Revolution. See Ukraine Indian institutions, 113. See also Sikhs
pure multinationalism, viii n3, 9–17, 20,
Pakistan: vs. India, 74–75, 75n54, 90–91, 34–35
109–10, 110n39; Jammu and Kashmir Puri, Harish, 93
and, 109–10; Kashmir and, 39n1, 110, putting together federalism, 26n30, 220n30
113–14, 114t, 115n43; Mizoram and, 102–
3; Mumbai terrorist attack and, 87; Quebec, 2, 26, 160, 263
Muslims in, 41, 54; vs. Nehru, 109–10;
trust in institutions of, 75n54. See also Rainwater, Lee, 273
Partition Requejo, Ferran, 12
Pallaver, Günther, 236 residual powers, 221–22, 261
Palshikar, Suhas, 87 Rezvani, David, 209–10
parliamentary systems, 20, 159, 208, 212, Riker, William, 11, 25n28, 257
304 index
robust multinationalism: alternative for- among, 48n20, 94; Hindus and, 83–85,
mulas for, x–xi; asymmetrical federal- 87, 92, 98; India and, 41, 84, 91–92, 94–
ism and, 262–63; defined, viii; 95, 98n10; Khalistan independence
democracy and, xi, 193, 257–58, 262–74; and, 39, 89, 95–97, 97t; massacre of,
parliamentarism and, 20; state-nations 74n52, 94–95; national identity among,
and, 17; vs. Ukraine, 174; in a unitary 61t, 62; and population in India, 41; vo-
state, 201; U.S. federalism and, 257–58, ter turnout among, 80, 80t
262–75. See also Belgium; Canada; Cat- Singh, Bhagat, 139t
alonia; India; Italy; Kashmir; multina- Sinhalese: on amnesty, 165, 166t; Bud-
tional federalism; pure multi- dhists among, x, 150, 155, 157, 161; and
nationalism; Spain; Sri Lanka; Tamil civil war with Tamils, x, 144–48, 152–55,
Nadu 157–61, 161n50, 165, 172; on discrimina-
Romania, 191–92 tion, 161–62, 162t; on federalism, 165;
Roper, Steven, 199n74 language issues of, 21, 150–53, 163–64,
Rose, Richard, 176, 180, 194–95 184; as majority, 144n3, 150, 153, 157; na-
Rudolph, Lloyd I., 118, 125 tionalism of, 151; nation building by, 161;
Russia. See Finland–Åland Island nation-state policies among, 150, 152–
federacy; Ukraine; U.S.S.R. 53, 157–58; peace sought by, 164–65,
Rustow, Dankwart A., 168, 244 166t; political leaders and parties of,
150–52, 155–57, 169; population of, 148;
Sardinia, 229, 231–32, 236–37 Sinhalese Only campaign in, 150–53,
Savarkar, Vinayak, 138, 139t 164; on Sri Lankan Army, 165; state pol-
SBY (Susilo Bambang Yudhoyono), 244– icies toward, 184n34. See also Sri Lanka
45, 245n104, 252n123 Skach, Cindy, 197
Scandinavia, 2, 11, 25n28, 190, 202, 210–11, Smeeding, Timothy M., 273
228, 234. See also Denmark; Norway; Smith, D. E., 153
Sweden Smyrl, Marc, 240
Scheduled Tribes. See India: marginal Sørensen, Axel, 226–27
groups in; Nagaland South India, 116–19, 123–24, 130. See also
scope values, xi, 36, 40–44, 114, 202, 228– Tamil Nadu; Tamils in India
29 South Tyrol, xi, 229–30, 230n60, 233–36,
SDSA (State of Democracy in South Asia) 241, 253, 256
study, 57n35, 61, 75, 161–64, 167 Soviet Union. See U.S.S.R.
secessionism, xi, 3, 8t, 10f, 20–21, 22n21, Spain: asymmetrical federalism in, 5–6, 9,
131n45, 265. See also cultural national- 37, 49, 165, 254, 261–63, 264f, 271; au-
ism; Flanders; Portugal; Punjab; South thoritarianism and, 65t; democracy in,
Tyrol; Tamil Nadu; Tamils in India 31, 36–38, 46n16, 63, 125, 196t, 263; fed-
secessionist nationalism, 18, 21–22, 22n21, eralism in, xii, 11, 49, 165, 257; identities
116–19, 124–27, 127n33, 129–34 in, 14, 29–31, 32t, 60–61; inequality of
semi-presidential systems, 17–21, 56, 159, representation in, 260t, 268t; multina-
169, 170t, 197–200, 199n74, 272t. See tional components in, 22–23, 29–30, 49;
also parliamentary systems; presidential national pride in, 58, 58t; nation-state
systems components in, 25–26, 28, 184; parlia-
Senanayake, D. S., 151, 156 mentary system in, 265, 271; population
Sicily, 229, 231–32, 236–37 of, 43; residual powers in, 261; as
Sikhs: Akali Dal party of, 85, 92–95, 95n9, robustly multinational, 2, 147, 264;
98, 98n10, 100; democracy supported Spanish Civil War in, 30; as state-
by, 69–71; economy of, 91n1; extremists nation, viii–ix, 11, 16, 25, 28, 36–38, 38t,
index 305
258; trust in institutions of, 38t, 74f; U.S. state-nation ideal types, viii, 7, 9, 13, 17,
federalism and, 258; veto in, 260, 271, 27–28, 37–38, 138. See also Ukraine
272t. See also Basque Country; Cata- state-nation overviews, 1–38, 89–114
lonia state-nations vs. nation states, vii–xvi, 1–
Sri Lanka: asymmetrical federalism in, 6, 38, 8t, 157–59, 173–74, 184–86, 188–90,
165, 170t, 172; career opportunities in, 254
129, 152–55; ceasefire in, 164–65; civil stateness, 10, 15n13, 109–15, 191, 242–43
war in, 5–6, 145–47, 158, 160, 161n53, State of Democracy in South Asia study.
165, 169–72, 172n64, 184, 191; colonial See SDSA
religions in, 153–55; democracy in, 5–6, Stepan, Alfred: Aceh and, 250n116; and
16–17, 141, 146, 148n14, 149, 167–71; fed- Burmese opposition, 266–67; democ-
eralism in, 5n5, 6, 170; Great Britain racy breakdown studied by, 82–83;
and, 153–54, 167; Hindus in, x, 144, 144– Finland–Åland Island federacy and,
45n3, 147–48, 152–55, 158, 161, 167, 167t, 246, 252n123; Gujarat model and, 84–
172; language issues in, 20–21, 150–53, 85; Hindi language issues and, 129;
157, 184, 193; Muslims in, 144–45n3, 162, identity conceptions of, 23; India/Kash-
162t, 164–65, 166t; nation-state policies mir conflict and, 110n39; Indonesia
in, 5–6, 150, 160, 162, 164, 166, 169, 172; peace process and, 250n118, 251n119;
peace in, 145–46, 160, 163–64, 166t, Linz’s collaborations with, xii–xiii, 4,
168–69; policy grammar in, 167–69; po- 23, 82–84, 110n39, 180; Mizo National
litical leaders and parties of, 149–53, Front (MNF) agreement and, 104n25,
155–60; political rights in, 169–72; presi- 105, 105n27; on SDSA board, 57n35; in
dential system in, 158–60; progressive Sri Lanka, 184n34; state-nation concept
nature of, 145–46; public opinion in, and, 4; Tamil Eelam restoration and,
161–68; robust multinationalism and, 159–60n46; Tamil principals inter-
148–49, 169, 174; semi-presidential sys- viewed by, 161n50, 164n54, 165, 165n56;
tem in, 170t; territorial divisions in, 147– Ukraine and, 179–80, 179n20, 183n32,
49; terrorist violence in, 146n11; tsunami 188n47; on veto players, 121. See also
in, 244; violence in, 145–47. See also Arel, Dominique; Djuli, Nur; Hale,
Ceylon; Tamils in Sri Lanka Henry E.; Jones, Sidney; Nurdin, A. R.
Sri Lankan/Ceylonese Buddhists: vs. affir- Sternberger, Dolf, 62
mative action, 163, 163t; on amnesty, Strachey, John, 50, 51n23
165; democracy supported by, 162–63, Subramanian, Narendra, 118–19, 129
163t, 168f; on equal rights, 167t; as ma- Suharto (Indonesian president), 242
jority, 144n3, 150, 153, 157, 162, 163t, 167; Susilo Bambang Yudhoyono (SBY), 244–
mobilization of, 151, 155–57; and myths 45, 245n104, 252n123
regarding Ceylon, 154; nationalism of, Sutarto, Endriartono, 243, 244n103
154–57; population of, 144, 162; state- Sweden, 3, 11, 74f, 190n51, 210–12, 213n16,
nation policies of, 162; state policies to- 216–17, 253–54, 272t
ward, 184, 184n34; state religion and, Switzerland: asymmetrical federalism in,
157, 170t; vs. Tamil/Hindus, x, 150, 155, 37; as coming together federation, 25,
157–58, 172; and trust in Sri Lankan in- 260; democracy in, 36–38, 46n16, 264f,
stitutions, 161–63, 163t, 167t 272t; inequality of representation in,
Stalin, Joseph, 186, 186n41 260t, 268t; national pride in, 58, 58n36,
state-nation characteristics, viii, 4, 7–8, 8t, 58t; nation-state components in, 184; re-
37–38, 173 ligion in, 68; as state-nation, viii, 4, 27–
state-nation enabling factors, 179–86 28, 38t, 58n36, 258; symmetrical federal-
state-nation goals, 191 ism in, 260–61, 264f; territorial mobili-
306 index
196t; independence of, 175–76, 179n20, tional pride in, 58, 58t; as nation-state,
181, 183; language issues in, 174, 176–78, viii, 36–38, 38t; political parties in, 81;
180, 183–93; minorities in, 176, 183–84, population of, 43; presidential system
192, 201; nationalists in, 182–86, 193; in, 262, 264–66, 271; Puerto Rico and,
nation-state policies and, 6, 6n7; Or- 209; religion in, 19, 46, 68; residual
ange Revolution in, 174n1, 179, 187–89, powers in, 261, 261n7; Senate, 259, 261;
193–94, 199; overview of, 173–200; separation of powers in, 262; Supreme
peace in, 191–93; polarization in, 187– Court of, 262; symmetrical federalism
88; political freedom in, 193; political in, 260–62, 264f; territorial mobilization
rights in, 174n1; religious issues in, 176; absent in, 2; trust in institutions of, 38t,
requirements for democracy in, 173–79; 74f, 76t; veto in, 121, 259–60, 262, 270–
vs. robust multinationalism, 174; Rukh 71, 272t; voter turnout in, 79, 80t
nationalist party in, 177–78, 182, 182n29; Uruguay, 63–65, 65n43, 65t
Russian irredentism and, x, 173, 178–79, U.S.S.R. (Soviet Union): disintegration of,
190; Russian language in, 188–93; Rus- 15; ethno-federal states of, 10, 17; Fin-
sians in, 6n7, 175–78, 180–82, 184–87, land and, 212, 216; putting together fed-
191, 194–95, 196t; state building in, 180– eralism in, 26n30; Stalin as positive
87; state-nation factors in, x, 179–86; figure in, 183, 186; stateness crisis in, 45,
trust in institutions of, 177, 195; U.S.S.R. 96. See also Ukraine
and, 174n1, 176–77, 193
unitary states: vs. asymmetrical federalism, Valle d’Aosta, 229–32, 236
17–18, 56, 170t, 203, 232; federacies and, Vanhanen, Tatu, 294
201–55; federalism and, 24–26, 25n26, Varshney, Ashutosh, xiii, 84–85
203, 275; independence from, 203–4; Verfassungspatriotismus (commitment to
nation-state policies and, x–xi, 3, 5, 8t, constitution), 12, 27, 62
173; separatist conditions in, 203; state- veto players, 121, 260, 271–73
nation policies and, x, 5–6, 8t; veto in, veto powers, 54, 121, 259–60, 262, 265, 271,
272t, 274t. See also Belgium; Denmark; 272t, 274t
federacies; Finland; France; Indonesia; voter turnout, 79–81, 95n8, 98
Japan; Luxembourg; Norway; Portugal;
Sri Lanka; Sweden; Ukraine Walloons/Wallonia, 33, 33t, 36, 263
United Kingdom, 5n5, 51n23, 160, 262n11, Walzer, Michael, 18–19
272t. See also Great Britain Way, Lucan, 174n1, 192, 192n56
United Nations, vii, vii n1, 90, 110, 226, Weber, Eugen, viii, 24
233, 235, 248, 253 we-feeling, 5, 7, 11, 13–15, 111, 161, 172,
United Nations Development Program, 186
xiii, 45t West, John, 218n24
United States: Åland Islands and, 212; Wheare, Kenneth C., 122n21, 257
amendment process in, 261, 271; Azores Wilson, A. Jeyaratnam, 159–60
and, 253; as coming together federation, Wilson, Andrew, 187n44
25–26, 46, 55, 120–21, 260; democracy Wolczuk, Kataryna, 178, 181
in, 36–38, 46n16, 264f, 272t; federal World Values Survey, viii, 37–38, 57–58,
model overview, xi–xii, 257–75; Green- 73–75, 77t, 140
land and, 226; House of Representa- Wriggins, Harold, 145–46, 152
tives, 259, 271; vs. India, 46–47, 55, 79–
81, 80t, 120–21, 257, 261, 264; inequality Yadav, Yogendra, ix, xii–xiii, 84–85, 87,
of representation in, xii, 259, 260t, 266, 134
268t; national identity in, 140, 141t; na- Yanukovych, Viktor, 176n7, 190, 194
308 index
Yeltsin, Boris, 182n30, 198, 198n73 Yushchenko, Viktor, 187–88, 190, 193–94,
Yudhoyono, Susilo Bambang (‘‘SBY’’), 199
244–45, 245n104, 252n123 Yusuf, Irwandi, 249–51
Yugoslavia, xii, 9–10, 14–17, 34–35, 45, 96,
144. See also Austria; Italy Zoramthanga (Mizo leader), 105