Journal of Ethnic and Migration Studies
Vol. 33, No. 6, August 2007, pp. 871 886
The Governance of Islam in Europe:
The Perils of Modelling
Veit Bader
Recently, we have seen a shift from research on the internal structure and culture of
Muslim religiosity to research on the way in which societies create opportunities for the
development of Islam, or oppose them, and more particularly to the political opportunity
structure and the institutionalised regimes and policies of governing Islam in Europe
from a neo-institutionalist perspective. This introduction to the special issue discusses the
concepts and perspectives of the governance and government of religious diversity,
critically analysing the inherent problems of constructing patterns or models of the
relationship between (organised) religions, societies, politics, nations and polities/states.
My analysis opts for fairly disaggregated frames for the purposes of rich descriptions of
cases, synchronic comparisons and diachronic changes, which are a precondition for
asking the relevant explanatory questions: Why what happened happened here and not
there? Why now and not then? This framework is used to critically assess the debate on
whether a European regime of religious governance in general, particularly with regard to
Islam, is emerging, and to introduce the contributions in this special issue that analyse
different aspects of governing Islam in Western Europe.
Keywords: Religious Diversity; Governing; Governance; Islam; Europe
The Muslim presence in Europe can only be called  new if we forget the lasting
impact of Al Andalus, the Ottoman Empire and European Colonialism. Yet Muslims
have gained in prominence, numbers and visibility, particularly in Western Europe,
after World War II due to post-colonial migration, the migration of guestworkers,
family reunification, refugees and asylum-seekers. Increasing numbers of Muslims
from different regions and countries, characterised by a broad variety of languages,
ethnic cultures and religious traditions in addition to differences in class-position,
Veit Bader is Professsor of Sociology in the Department of Political and Socio-Cultural Sciences and of Social
and Political Philosophy in the Department of Philosophy at the University of Amsterdam. Correspondence to:
Prof. V.M. Bader, Dept of Philosophy, University of Amsterdam, Nieuwe Doelenstraat 15, NL 1012 CP
Amsterdam, The Netherlands. E-mail: [e-mail]
ISSN 1369-183X print/ISSN 1469-9451 online/07/0600871-16 # 2007 Taylor & Francis
DOI: 10.1080/13691830701432723
872 V. Bader
age, citizenship status etc., have tried to live their religious practices as minorities
under conditions of entrenched but widely diverging regimes of governance of
religious diversity. Moreover they have increasingly lived in a transnational context*
in the triple meaning of demographic migration, of diasporas, and of increasingly
global public spaces (Bowen 2004: 891).
On the one hand, these partly-new conditions had a considerable impact on
Muslims everyday religiosity and on the doctrinal articulations of Islam; on the ways
in which they raised common demands or claims; on the ways in which they
identified, associated and acted collectively in order to change existing administrative
practices, legal rules, predominant policies, and even the institutional structure and
collective identities of the receiving societies and states. On the other hand, Muslims
presence, claims and collective actions also had an impact on institutionalised
regimes of government and governance in receiving societies, both individual states
and the European Union; on policies of assimilation or accommodation of religious
diversity; and on predominant ideologies, images and collective identities.
After some delay, studies of these dynamic interactions have entered into academic
research agendas, initially slowly. In the last ten years or so, however, the number of
articles, books and reports on nearly all aspects of the Muslim presence in Europe has
literally exploded,1 and the introduction to this special issue is certainly not the place
for even the shortest state-of-the-art report. Researchers from different disciplines,
different research traditions embedded in different  national academic and political
cultures, and working increasingly in close cooperation between internationally-
oriented and interdisciplinary research centres, have produced a widely hetero-
geneous body of literature on changing Islamic religiosity (mainly by anthropologists
and sociologists of religion, theologians, philosophers and normative legal theorists)
and on changes in the governance of Islam (mainly by political scientists,
institutionally-oriented sociologists and comparative legal theorists). Broadly speak-
ing, we can see a shift from research of the internal structure and culture of Muslim
religiosity*the formation of their associations and organisations, the development
of Muslim identities (if any)*to the  way in which societies create opportunities for
the development of Islam, or oppose them (Buijs and Rath 2003: 9), in other words,
to the external opportunity structure generally. This latter includes the political
opportunity structure and the institutionalised regimes and policies of governing
Islam. We also increasingly note the  dialectical (Césari 2005) or dynamic interaction
of internal and external aspects.
Even in the narrow perspective of the Governance of Islam it would be futile to
discuss the most important studies of specific issues like Islamic instruction in
governmental schools or religious dress codes, or the dynamics of accommodation in
specific units of governance or government*like firms, municipalities, nation-states
or the EU. Instead, I focus on theoretical and methodological problems in this
promising but still under-developed governance approach to the study of Islam(s) in
Europe (Islam en Europe) and the contested emergence of a  European Islam (Islam
d Europe). First, I explain the concepts of governance and government that emerge
Journal of Ethnic and Migration Studies 873
from recent institutionalist approaches in the social sciences. In the next section I
address inherent problems of constructing models of regimes of governance of
religions. Then I briefly discuss whether something like a European Regime of
Religious Governance is likely to emerge. In the final section I elaborate the specific
focus of this special issue on Governing Islam in Western Europe, and introduce the
articles in this issue.2
Governance and the Government of Religious Diversity
Governance is an increasingly fashionable concept nowadays but it has not yet been
applied very much to problems of ethnic and national minorities as reflected in the
literatures on multiculturalism in general and on religious minorities in particular.
The governance of religious diversity has also been widely neglected in the sociology
of religion, which traditionally focuses on describing and explaining the diversity of
religious beliefs and practices.
Governance may be a fashionable concept but it is often used in a vague and
unspecified way (Héritier 2002; Pierre and Peters 2000; Treib et al. 2005). Hence I
start by briefly summarising the conceptual distinctions between patterns or regimes
of governance and of government as used in comparative institutional studies in the
social sciences. Structured patterns (or  formations or  configurations ) comprise
relevant interactions between economic, social, cultural (including ethno-national),
political, legal, judicial, administrative and religious relations in all their diversity; to
put it more traditionally, the relationships between economy, society, culture, politics,
nation, state and (organised) religions. Studies of society religion patterns have to
analyse all mechanisms of action-coordination*markets, networks, associations,
communities, private and public hierarchies (Hollingsworth and Boyer 1997 )*and
all relevant (coalitions of) actors, their organisations and strategies, including
governments at all levels.
The focus of governance is narrower: regulation or steering, guidance by a variety of
means, not only by rules. It includes only those mechanisms of action-coordination
that provide intentional capacities to regulate, including co- and self-regulation.
Markets, on the one hand, are important mechanisms for co-ordinating actions but
they do so by an  invisible hand. They have no regulatory capacity and therefore
should not count as modes of governance. The perspective of governance, on the
other hand, is much broader than that of government, which focuses on one
(internally highly diversified) actor*the state*and on action-coordination by
 public hierarchy , by rules, particularly law and law-like regulations. Governance,
then, includes more actors and more modes of coordination in the perspective of
regulation and is well-suited to describe and explain recent  shifts from government
to governance . Yet it excludes some broad issues beyond governance: not all co-
ordination is governance, only co-ordination by  policies in a very broad sense.
Governance in general can best be understood by discussing it along two
axes of regulation, distinguishing internal and external governance and democratic
874 V. Bader
(bottom-up) and hierarchical (top-down) governance. Both internal and external
governance can be top-down or bottom-up.
Let me apply these distinctions to explain three points about religious governance.
First, the relatively free competition of all types of religion in  God s Biggest
Supermarket may be crucial for the opportunities for divergent religions and for
overall religious diversity, but this is not in itself a mode of governance. It is an
invisible-hand mechanism that is, in analogy to rules of contract and property,
regulated everywhere at least implicitly (e.g. by freedom of religious conscience and
other religious freedoms) and which has to be regulated by customs, conventions or
laws, by self-regulation within and among competing religions, by co-regulation and
by public bodies.
Second, religious governance refers to the internal and external regulation of
religious diversity and to their dynamic interaction. Internal governance by the
respective religious communities themselves*here understood as those groups
which share certain religious beliefs and practices, however contested the boundaries
and the beliefs and practices may be*includes  self -regulation by religious laws and
customs (ecclesiastical law and nomos) of many aspects of life from the cradle to the
grave. The rules and their interpretation and application can either be more
autocratic and hierarchical*e.g. formally organised, autocratic church hierarchies
and religious elites*or more democratic and bottom-up*e.g. by religious
congregations and democratically elected religious organisations/leaders, or indeed
by more informal networks and associations of believers, or religious counter-elites
like dissenting theologians, leaders of religious political parties and communal
business elites.3 Internal governance of religious diversity should not be called
 management (Bouma 1999) because management includes hierarchy and top-down
competencies usually confined to either private or public hierarchies. Internal
governance of religious communities clearly varies widely between religions (e.g.
Catholicism approaching the autocratic pole, radical Protestantism the democratic
one, and Islam the less-organised one) and, obviously, also in the historical
development of specific religions.
Third, external governance of religious diversity also includes more voluntary and
democratic forms of self-regulation by interfaith networks, associations and
ecumenical organisations, which are often neglected and cannot appropriately be
captured by  management of religious diversity . The governance perspective also
stimulates analysis of the regulations of religious diversity inside so-called semi-
public and private organisations by more or less autocratic or democratic forms of
 corporate governance , as well as the analysis of their impact on the ways in which
religious practices (particularly of religious minorities) are accommodated or not.
Most attention has been focused on external governance of religious diversity by
governments or public hierarchies*in other words what polities (on different levels,
from local to supra-state), legislations, administrations, jurisdictions and the different
departments of government do to religions, particularly by law or law-like rules.
Comparative studies of the divergent ways in which religions, particularly Islam,
Journal of Ethnic and Migration Studies 875
are governed in Europe have been initiated by legal scholars (e.g. Ferrari 1996;
Ferrari and Bradney 2000; Shahid and van Koningsveld 2002). Much less research has
been done on what states (e.g. courts, administrations) actually do, on how these
diverse and changing practices relate to the country-specific predominant legal
models or ideological policy-models, and on how different religious minorities
respond to those policies, which include not only legal opportunities and threats but
a broad spectrum of other resources (money, expertise, networks etc.) and policy
options (incentives and persuasions).
Compared with the traditional focus on the government of religious diversity, the
perspective of governance promises two important advantages. First, if governments
follow policies of deregulation and  privatisation also regarding religions, the
governance perspective helps to prevent a dichotomous view of  either state or
market and enables analysis of the huge variety of public but non-state actors and of
semi-public and private actors which play such a big role in regulating religious
diversity anyway and are gaining in importance by such  shifts from government to
governance .4 Second, it is more appropriate for the study of religious regimes like
American denominationalism that are characterised by fairly thin legal regulations
and little  official cooperation between state and religions.
The Perils of Modelling: Path Dependency and Change
Societal patterns and regimes of religious governance together define the general
external opportunity structure that, according to an institutionalist approach,
contains the most important variables for explaining variations in Muslim religiosity
and in the accommodation or non-accommodation of Muslim demands in Europe.
Researchers looking for help by sociologists of religion only detect a lot of
disagreement and competing attempts to construct such models. Here I briefly
discuss the perils of all modelling but I also elucidate why we cannot do without such
models.5
. Patterns*e.g. the American, English, Scandinavian, Mixed, Latin, Right Statist,
Left Statist, and Nationalist patterns distinguished by Martin (1978)* require
some degree of stability in a diachronic, historical perspective.  Nested structures
exclude other institutional options, and  path dependency , in spite of its critical
edge against  necessitarian evolution, excludes other paths (Crouch and Farrell
2004). Yet, path dependency has to allow for modifications and even breaks of
patterns: history is not destiny. In constructing patterns, one has to avoid the
danger of presenting them as more stable over time than they actually are. Area or
state patterns change over time or are even revolutionised.
. Societies may be better described as patchworks or bricolages than as orders with a
very high degree of systemic integration. Yet the fact that they are only  loosely
connected does not exclude attempts to detect weaker forms of  structural
interdependency or  institutional complementarity , and of vicious or beneficial
876 V. Bader
cycles. If one constructs internally fairly homogenous patterns, one should be
aware of the danger, inherent in comparative methods, of neglecting important
internal heterogeneity: states in the same area-cluster differ from each other,
regions within states differ from each other, and so on.
. Informative models should be more than historical, singular, one-case narratives
(Minkenberg 2003a: 120). Two methods to construct comparative models are
available: inductive generalisation and deductive theoretical construction. Induc-
tive generalisations are usually more sensitive to the complexity of (overlapping)
dimensions but often lead to a proliferation of more and more patterns. Deductive
theoretical constructions are more parsimonious, using only one or two
dimensions, resulting in fewer*usually two or by cross-tabulation four*
patterns. There seems to be a clear trade-off in this regard between parsimony
and consistency on the one hand, and historical and comparative sensitivity, on
the other: the more patterns or types, the higher the historical sensitivity and the
empirical  fit with cases, but the more complex and eventually useless these
models become for comparative research.6 Parsimonious, consistent models,
however, seem to lack minimally required historical sensitivity and empirical fit.
Patterns, after all, have to capture, condense and articulate crucial structures and
traits that really characterise areas, polities or historical epochs. Obviously, there is
no easy solution here but the construction of theoretically guided inductive
models may be the best way forward.
. The (territorial) units can be areas with religio-cultural and/or legal-political
characteristics, and may be coextensive with empires but normally contain more
than one polity and often even cut across polities (regimes of religious governance
are normally not confined to states). Polities, particularly states, however, are
important as units of definitive collective political decision-making (regimes of
government of religious diversity).
. Models containing many dimensions*e.g. constitutional, legal, administrative,
judicial, social, political and cultural*are historically and empirically more
informative than models focusing on one dimension only*usually the constitu-
tional/legal relationship between  State and Church *but they are also usually less
consistent because all these dimensions, though interacting, generally do not
change simultaneously or in the same direction.
. The operationalisation of the respective dimensions in indicators and criteria may
contribute to make models, particularly multi-dimensional ones, over-complex,
making testing even more difficult.
. For each relevant dimension (and each indicator) one can choose between a
dichotomous (allowing two poles or extremes only) and a gradualist strategy
(allowing for several options on a scale between the extremes). Again, there seems
to be a trade-off: dichotomous modelling is consistent and parsimonious at the
price of neglecting important options, gradualist modelling is more murky but
open to relevant*descriptive and prescriptive (normative and practical-
political)*distinctions.
Journal of Ethnic and Migration Studies 877
. Finally, models can be constructed for descriptive and/or explanatory purposes or
for normative purposes (both evaluative and prescriptive ones). Obviously, the
line between the two is only analytically sharp and it helps a lot if normative, often
implicit assumptions in modelling are made explicit.
In addition, one should distinguish as clearly as possible between (1) explaining the
emergence, stabilisation and reproduction of distinct patterns (here patterns are
explananda or  dependent variables); (2) changes and/or break of patterns (we
should not expect that which the same causes explain the emergence and stabilisation
could also explain the reproduction, change and demise of patterns); and (3)
explanations of diversities in religious politics in general (Minkenberg 2003b, 2004),
and the incorporation of (new) religious minorities in particular (e.g. Fetzer and
Soper 2005; Koenig 2003, 2004, 2005), in which patterns may serve as explanans or
 independent variable *obviously in combination with other causes (Humphreys
1989).
In my view, we need fairly disaggregated models of the relationship between
religions, societies and polities/states. Such models should contain all relevant
dimensions or  variables both for descriptive and for explanatory purposes. These
models are theoretically guided: they have to explain questions like which dimensions
are  relevant , and which are  independent and  dependent variables. These
disaggregated conceptual, pre-theoretical frames could be used in a direct way for
rich descriptions of cases, for informed synchronic comparisons and the analysis of
diachronic changes. Moreover, they allow asking the relevant explanatory questions:
why what happened happened here and not there, why now and not then?
If we zoom in now on regimes of government of religious diversity relevant for our
discussion of  governing Islam in Europe , we first need to disaggregate government.
Actual state policies with regard to religions have to be analysed along four different
axes:
. Constitutional, legal and administrative regulation, and political and cultural
interference. Constitutional regulation (if any) articulates religious freedoms and
the specific forms of establishment or disestablishment of churches or religions.
Legal regulation is understood as the traditional combination of legislation and
jurisdiction, often in a  multi-level system (Galanter 1966); it may also imply
normative legal pluralism, e.g. recognising private religious law and jurisdiction.
Legislative regulation may be fairly general and limited or very specified and
extensive, depending on the legal traditions of countries and on the actual degree
of states regulatory and policing powers*which have increased dramatically with
modern welfare states (Ferrari 2002; Monsma and Soper 1997). Under conditions
of religious diversity in modern welfare states the administrative interference of the
state is inevitably extensive and varied, and the degree of administrative discretion
is fairly high (guided by general norms like  public order ,  equity ), particularly if
legislation is general or laws are supplemented by general executive orders. Even if
878 V. Bader
guided by the idea of strict or formal neutrality, legislation and administration
under such conditions cannot be neutral in their effects (Laycock 1997), and
jurisdiction by administrative or constitutional courts would inevitably unfairly
restrict freedom of religious exercise if guided by such a meta-legal principle
(Monsma and Soper 1997). Legislation, administration and jurisdiction inevitably
interfere politically in public debates regarding state religion relationships and
also may explicitly attempt to exert some cultural influence. One consequence of a
differentiation along such lines is that legal recognition should not be confused
with political and administrative recognition.  Formal legal recognition of newly
organised religions is not a necessary precondition of actual recognition; it may
even impede de facto administrative recognition, as the cases of Germany or
Belgium demonstrate. Another consequence is that legal guarantees of individual
or collective religious freedoms do not imply actual protection (as the predicament
of Catholics in the nineteenth-century US or of Muslims in Britain today clearly
shows).
. The aims of states vary widely. States may attempt to suppress and prosecute all
individual or collective religious beliefs and practices or specific (e.g. new
minority) religions, they may aim at tolerant negligence, at formal or actual
protection of religious freedoms of all or of specific religions, at relational
neutrality, or at privileges for all or some (established) religions. States can choose
(a specific mix of) different strategies in order to achieve their aims to suppress or
protect (all or specific) religions, treat them equally or positively facilitate and
privilege (all or specific) religions (Rath et al. 1996: 245). They may control,
concede, separate; they may persuade, reward or punish; they may apply direct or
indirect and reactive or proactive strategies (Bader 1991). In addition, states can
use a variety of means to realise their aims and strategies. The specific means of
states are laws and jurisdiction (monopolies of legislation and jurisdiction) and the
(threat of) enforcement (monopoly of legal violence). States, however, can use
many other resources in order to influence religions: material resources (land,
buildings etc.), qualification of state personnel, money, information, political
legitimacy.
. The state is not a homogeneous, monolithic actor: its different branches
(legislative, executive etc.) as well as its different executive departments do not
necessarily follow the same aims, strategies and policies. Justice and Home Office
departments differ thoroughly from departments of education, health-care,
housing and urban planning, cultural and social affairs etc., particularly in
situations of declared security threats, and the same holds, obviously, for the
different levels of state organisation: federal, state and local administration. Local
administrations are facing urgent and legitimate demands of religious immigrants
more directly and are generally more accommodationist than central government.
. State laws and policies do not only, directly or indirectly, regulate highly sensitive
religious issues like family, sex and gender matters, abortion, euthanasia, genetic
engineering, and education, but also many other policy-issues like employment,
Journal of Ethnic and Migration Studies 879
business, housing, health-care and welfare, all of which may be crucially important
for religious minorities (Minkenberg 2003b).
Why, then, should we construct models of more aggregated patterns at all? First and
foremost, the complexity of the relationships, dimensions and indicators is
overwhelming and the task of modelling and theorising is not to drown us in
unstructured complexity but to help reduce this complexity in a controlled and
reflective way. Historical and comparative research is in need of explicitly constructed
typologies; even case-studies use models implicitly. If one constructs less complex,
theoretically guided models, a disaggregated  proto-theoretical frame necessitates
answering the relevant questions: which are the relevant or important dimensions to
be included and why is this the case?
In order to test the validity and reliability of these constructs, we need to break
them down into components, and to see whether the various elements subsumed
under a specific label hang together as well empirically. A preliminary analysis has
distinguished seven major components of church state relations. . . . (O)nly after
decomposing large concepts in this or in other alternative ways, and only after
identifying the empirical indicators, may one hope to find answers (Enyedi 2003:
225).
Secondly, even if societies and, more specifically, the religions states politics
relationship, are not fully systemically integrated, they show some minimal internal
coherence, and the construction of patterns has to demonstrate these linkages,
mechanisms and trends. This is more urgent for researchers who are not content with
multivariate statistical analysis and aim at some  causal ,  structural or  functional
explanations (Bhaskar 1975; Humphreys 1989).
Thirdly, fairly aggregated models of the religions state policies relationship are
actually constructed and used by state and policy-makers in their design and/or
legitimation of institutions and policies, as well as by political philosophers, political
and legal theorists, judges, administrators, political parties, journalists, NGOs etc.
These institutional and policy models of the desirable or acceptable relationship
between states, policies and religions are explicitly normative. In spite of the variety of
policies across fields and levels, state policies are guided, or at least legitimised, by
very general and rough models of desirable end-states and by acceptable ways to
achieve these aims. Competing models are important for critics and opponents.
In my view, it is a fair guess to assume that the highest degree of aggregation and
of consistency and coherence of policies will be found in ex ante legitimations
and designs. Consistency decreases when it comes to more detailed laws,
policy documents and actual policies in different fields and on lower levels.
The task of social science is to describe and explain these predominant and
oppositional normative institutional and policy models, their actual impact on
policies, and their effects. This includes the analysis of the power relations between
the different (coalitions of) actors constructing and using such models or  discursive
880 V. Bader
frames *politicians, judges, philosophers and social scientists as public intellectuals,
journalists etc.*and their impact on public discourse. In policy-evaluation studies,
the sharp line between descriptive and prescriptive analysis already becomes blurred,
though they are usually guided by the stated aims of policy-makers (Bader and
Engelen 2003: 382 8). All those who design alternative institutions and policies or
want to defend or modify existing ones, have to spell out their own normative
principles (and political philosophers and theorists are considered to be normative
specialists in this regard).
Fourthly, the gap between these predominant normative models of appropriate
institutions and policies and  what is going on on the ground , the actual institutions
and policies, is expectably huge in all countries but particularly in countries like
France or the US, where ideal models of  strict separation of state and religions are
paramount (see Bader 2007). It should be evident that normative models do not
determine institutions and policies, but the same is also true for empirically
institutionalised patterns of state/politics and (organised) religions. In explaining
differences in the state accommodation of religious practices by Muslim immigrants
in Germany, the UK and France, Fetzer and Soper (2005) convincingly demonstrate
that  Church State theory is more powerful than competing theoretical approaches
like Resource Mobilisation, Political Opportunity Structure, and Political Ideology. In
their article in this special issue, Fetzer and Soper make clear that State Church
relations do not  determine but  shape accommodation policies. Indeed, established
or institutionalised patterns, like principles and rights (of religious freedoms, for
example) have been and have to be continuously re-interpreted and re-framed, and
framing depends on competing discourses of incorporation, on discourse coalitions
and power relations, and on crucial events (see Koenig, this issue). The predominant
(interpretation) of State Church relations in France is historically changing and it is,
at any given time, not sufficient to explain, for example, local variations in the
construction of mosques in France (see Maussen, this issue).
An Emerging European Regime of Religious Governance?
Constitutional, legal, administrative and political regulations of religions in European
states vary widely (Bader 2007). Constitutional laws range from stronger or weaker
versions of establishment of one or two state churches to dis- and non-establishment.
Constitutional reality diverges more or less drastically from formal regulations.
Among scholars it is common practice to distinguish three recent patterns: countries
with a state or national church, strict separation countries, and countries with more
or less dense selective cooperation between governments and (organised) religions.
The legal status of religions ranges from no special treatment through minimal to
wide and deep legal recognition and regulation. Though all states have to guarantee
the autonomy of (organised) religions from state regulation and interference in their
internal affairs, rules and practices differ considerably. All liberal states finance
religions at least indirectly by means of tax exemptions, but some also by general
Journal of Ethnic and Migration Studies 881
subsidies or vouchers. Some states finance religions directly by paying salaries or
other costs, by religion-specific subsidies and even modern varieties of  church taxes .
Most European states also finance (and regulate) faith-based educational institutions
in different ways, and all finance (and regulate) faith-based care and social service
institutions. Regulations and practices of religious instruction and the position of
teachers in governmental schools also differ widely and the same holds for chaplains,
for admission rules for religious teachers and ministers of religions (Kraler, this
issue).
This bewildering variety of rules and practices, which increases drastically if one
takes historical changes and local and provincial differences within states into
account, means that we have good reasons to expect that these overlapping and
intersecting patterns change, but never simultaneously or always in the same
direction. Actually, we see a huge diversity of contradictory constitutional, legal,
administrative and cultural arrangements and policies within states, all sharing the
same principles of liberal democracy. This institutional diversity does not seem to
converge into one optimal or even predominant regime of religious governance,
neither globally7 nor for all Western states, and not even for the member-states of the
European Union.
A common European regime of religious governance, broadly speaking, would
require not only the emergence of a common regime of  governing religions or a
 common West-European pattern of church and state relations (Ferrari 2002: 7), but
also the convergence of associational, network, communal-cultural and (semi-)pri-
vate modes of governance of religious diversity. Some authors claim that we are
witnessing the emergence of a European regime of government of religious diversity
showing three main features:  the right to religious liberty ,  the religious incompe-
tence of the state and the autonomy of religious groups , and  selective co-operation
of states and religious groups (Ferrari 2002: 7 11).8 I confine myself to three critical
objections.
First, the right to religious liberty, the incompetence of the state and the autonomy
of religions are, of course, important minimalist moral and legal constraints. This
 coercive isomorphism by way of  legal transnationalism (Koenig 2003) through
binding international covenants, regional human-rights regimes and also soft law9 are
clearly important but do not lead  to full institutional isomorphism. Rather, it should
be conceived in terms of processes of normative pressure and imitation amounting to
a successive diffusion of cognitive and normative schemata (Koenig 2005: 230). Still,
two caveats are in place. These principles and rights are interpreted, weighed and
applied in widely divergent ways in  national traditions and jurisdictions within the
EU (Bowen, this issue; Fetzer and Soper 2005; Monsma and Soper 1997). In addition,
the emergence and transnational diffusion of a multi-cultural citizenship model and
of pluralistic modes of organisational incorporation may not be as strong,
uncontested and irreversible as Koenig and others assume (see Bader 2005).
Second, it is fiercely contested whether something like the institutional model of
 selective cooperation really emerges because at least some member-states (such as
882 V. Bader
France) massively oppose this option and the EU has no regulatory competence in
this regard to  harmonise rules and practices. In addition, selective cooperation
covers a broad variety of institutionalised ways and means of cooperation between
state and organised religions.10
Third, the robustness or staying power of path-dependent and particularly of
institutional or structural differences seems greater, and the effectiveness of
 isomorphic change through the three causal mechanisms of enforcement, imitation
and normative pressure seems weaker than authors like Koenig (2004) have assumed.
Claims that the  global transformation of the classical models of nation states subjects
the institutional logics of religion-policies (. . .) to a fundamental change (Koenig
2004: 87) leading, amongst other things, to an increasing convergence of Western
European Islam-politics in the 1990s, are not corroborated by actual developments.
The logic and the relative impact of path-dependent divergent institutional
arrangements and of convergent trends in general, and within the EU in particular,
are subjects of heated debate, and some of the articles in this issue significantly
contribute to a further clarification of the inherent difficulties.
Governing Islam in Western Europe
The articles in this issue can be seen as first steps towards a more demanding
comparative research programme on the Governance of Islam in Europe. They share
an institutionalist approach, broadly speaking, giving due weight to differences in
external opportunity structures, particularly political opportunity structures, in a
governance perspective. Yet, as a first step, they mainly focus on  Governing Islam .
They also share a comparative multi-level and multi-field approach and may help to
increase our knowledge of under-researched aspects of processes of incorporation of
Muslims in Europe in the following four ways.
Firstly, Minkenberg takes up the comparison of governing Islam with other
religions, and puts it into a broader perspective, discussing the chequered relationship
between democracy and religion, particularly the  pasts, paths, and presence of
Christianity and democracy, and of Islam and democracy.
Secondly, some of the papers explicitly address the issue of whether a
Europeanisation of governance of religious diversity can be expected and detected
(Koenig s multi-level analysis of the EU).
Thirdly, the articles demonstrate the persistent impact of the diversity of  national
patterns on policies of religious accommodation in EU member-states. Soper and
Fetzer recognise that national regimes are under pressure (from common problems,
European regulations) and are also controversial within countries, but they defend a
fairly strong thesis of path-dependent national  structuration of policies in their
comparative analysis of Britain, France and Germany that also shapes the three
possible futures of Islam in Western Europe. Kraler analyses the divergent special
admission rules for ministers of religion in nine European countries and Canada,
focusing on admission policies vis-Ä…-vis Muslim ministers in the context of recent
Journal of Ethnic and Migration Studies 883
desires to control and monitor Muslim leadership against the wider background of
attempts to  Europeanise Islam. Bergeaud-Blackler describes divergences of integrat-
ing Islamic ritual slaughter into the legal systems of Western European countries.
Finally, by focusing on historical changes and internal varieties of national regimes
of governing religious and cultural diversity, particularly of the regulation of Islamic
presence in one member-state, France, Maussen shows how heterogeneously  the
French model has been interpreted and applied.
In his concluding remarks to the special issue, John Bowen highlights the inherent
problems in country and issue comparisons and their limitations. Instead of
introducing new dimensions of disaggregation of national models along the line
argued in Maussen s paper, he asks  about other properties of these national models
and how we might look comparatively at the very problematic character of the
national model itself , focusing on models as cultural resources for political action on
which actors are able to draw at particular moments rather than as explanatory
models. A short history of France s laïcité vividly demonstrates that its ubiquity
derives not from its provision of a prescriptive framework for the relations of
religions and state in France, but because it actually conceals a great deal of conflict,
disagreement, and general muddle about what those relations are or should be. In
addition, Bowen cautiously reminds us that the increasing difficulties faced by social
scientists in constructing models of governance of Islam in Europe will provide a
useful index of Muslims own success.
As first steps towards a more comprehensive analysis of the governance of Islam in
Europe, the articles in this special issue clearly share some limitations. First, in order
to get a fuller picture of  governing Islam we urgently need more detailed studies on
what  governments on all levels actually, and not only legally, have done and  do to
Muslims . Second, we should investigate the two fairly neglected areas of governance
indicated at the beginning of this article: research on internal governance (both top-
down but maybe even more importantly also bottom-up), and research on the
corporate governance of Islam in a comparative perspective. Third, even a fuller
analysis of governance (what governments, other actors, and Muslim organisations
and leaders  do to Muslims ) is clearly only part of the whole story of incorporation
processes. This is because polities are not all-powerful, policies have unintended,
counterproductive effects, and a lot depends upon  what Muslims do to states and to
other actors, including their own organisations and leaders. Still, the  governance
perspective highlights the enormous, ongoing power asymmetries in these
 dialectical encounters, deliberations, negotiations and conflicts.
Notes
[1] See Buijs and Rath (2003). See also the earlier JEMS special issues on  Islam,
Transnationalism, and the Public Sphere in Western Europe (Grillo and Soares 2004) and
on  Mosque Conflicts in European Cities (Césari 2005).
[2] This special issue originated in a workshop held at the University of Amsterdam in May 2005
as part of the ongoing activities of the EU Network of Excellence on  International
884 V. Bader
Migration, Integration and Social Cohesion in Europe , specifically the IMISCOE Research
Cluster B6.
[3] See Kalyvas (1996); also Mazower (2005) on the internal governance of the Jewish
community in Salonica.
[4] For a concise summary of recent developments see Kersbergen and Waarden (2001);
Schmitter (2000).
[5] For a more extensive discussion of different approaches see Bader (2007). All these problems
are more or less extensively discussed in attempts to model the variety of institutions and
policies in other areas. For the  types or  models of capitalism debate, see Amable (2003);
Crouch and Streeck (1997); Hollingsworth and Boyer (1997); Kitschelt et al. (1999); Whitley
(2000). For varieties of  welfare states see Esping-Andersen (1990); Pierson (2000); for types
of democracies, Lijphart (1984); for regimes of incorporation of minorities, Bader (1997).
[6] The refinement of types and subtypes eventually leads to models presenting each country as a
separate type. In addition, one would still have to answer the question: Why states and not
sub-national regions?
[7] Contrary to the claim raised by Bouma (1999) and many others. At this level,  global
government sensu strictu ( coercive isomorphism ) is absent, and convergence would have to
result from a loosely structured  world-polity , from  mimetic isomorphism (imitation, best
practices), from  normative isomorphism (Koenig 2003), and/or from other modes of
governance.
[8] Massignon stresses the third aspect, the emergence of a  concordatory Europe that is
 pluralist with structural limits (2003: 5), situated  halfway between the pluralist inter-
denominational American model and the classic European model of a hierarchy of
recognised religions . In addition, Enyedi points to  converging tendencies between Europe
and the US . . . Partly as a result of the policies pursued by the Bush administration,
European-style state support for churches has attracted considerable interest. The irony is
that, while in the USA churches and politicians have begun to embrace the idea of closer
cooperation between church and state, in Europe, the principle of separation finds growing
support among religious sectors (Enyedi 2003: 219).
[9] Like the  Declaration on the Elimination of all Forms of Intolerance and Discrimination
Based on Religion and Belief (1981), the UN Convention on the Protection of the Rights of
Migrant Workers (1990), the UN Declaration of the Rights of Persons Belonging to
Minorities (1992) and the various activities of the Council of Europe (European
Commission Against Racism and Intolerance (ECRI) and of the Organization for Security
and Co-Operation in Europe (OSCE).
[10] I agree that  governments increasingly cooperate with organised religious bodies in many
institutional fields including education, welfare provision , and maybe even in  legislation and
jurisdiction (Koenig 2005: 230; see also Bader 2007) but the way in which they do remains
importantly shaped by path-dependent patterns. Compared to his 2003 dissertation and
2004 article, Koenig has highlighted more clearly the impact of  historical path dependencies
of institutional arrangements in his papers published in 2005 and in this special issue.
References
Amable, B. (2003) The Diversity of Modern Capitalism. Oxford: Oxford University Press.
Bader, V. (1991) Kollektives Handeln. Opladen: Leske & Budrich.
Bader, V. (1997)  Incorporation of ethnic or national minorities. Concepts, dimensions, fields, and
types . Online at http://home.hum.uva.nl/oz/baderv.
Bader, V. (2005)  Dilemmas of multiculturalism , Canadian Diversity/Diversité Canadienne, 4(1):
86 9.
Journal of Ethnic and Migration Studies 885
Bader, V. (2007) Secularism or Democracy? Associational Governance of Religious Diversity.
Amsterdam: University of Amsterdam Press.
Bader, V. and Engelen, E. (2003)  Taking pluralism seriously: arguing for an institutional turn in
political philosophy , Philosophy and Social Criticism, 29(4): 375 406.
Bhashkar, R. (1975) A Realist Theory of Science. Leeds: Leeds Books.
Bouma, G. (1999)  From hegemony to pluralism: managing religious diversity in modernity and
post-modernity , Australian Religion Studies Review, 12(2): 7 27.
Bowen, J. (2004)  Beyond migration: Islam as a transnational public space , Journal of Ethnic and
Migration Studies, 30(5): 879 94.
Buijs, J. and Rath, J. (2003) Muslims in Europe: The State of Reasearch. New York: Russell Sage
Foundation; Amsterdam: IMES.
Césari, J. (2005)  Introduction , special issue on Mosque Conflicts in European Cities, Journal of
Ethnic and Migration Studies, 31(6): 1015 24.
Crouch, C. and Streeck, W. (1997) Political Economy of Modern Capitalism. London: Sage.
Crouch, C. and Farrell, H. (2004)  Breaking the path of institutional development? , Rationality and
Society, 16(1): 5 43.
Enyedi, Z. (2003)  Conclusion: emerging issues in the study of Church State relations , in Madeley,
J. and Enyedi, Z. (eds) Church and State in Contemporary Europe. London: Frank Cass,
218 32.
Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity.
Ferrari, S. (2002)  Islam and the Western European model of Church and State relations , in Shadid,
W.A.R. and van Koningsveld, P. (eds) Religious Freedom and the Neutrality of the State: The
Position of Islam in the European Union. Leuven: Peeters, 6 19.
Ferrari, S. (ed.) (1996) L Islam in Europa. Bologna: Il Mulino.
Ferrari, S. and Bradney, A. (eds) (2000) Islam and European Legal Systems. Aldershot: Ashgate.
Fetzer, J. and Soper, C. (2005) Muslims and the State in Britain, France and Germany. Cambridge:
Cambridge University Press.
Galanter, M. (1966)  Religious freedoms in the United States: a turning point , Wisconsin Law
Review, Spring: 217 96.
Grillo, R. and Soares, B.F. (eds) (2004)  Islam, transnationalism and the public sphere in Western
Europe . Special issue of the Journal of Ethnic and Migration Studies, 30(5): 861 1031.
Héritier, A. (ed.) (2002) Common Goods. Oxford: Rowman & Littlefield.
Hollingsworth, R. and Boyer, R. (eds) (1997) Contemporary Capitalism. Cambridge: Cambridge
University Press.
Humphreys, P. (1989) The Chances of Explanation. Princeton: Princeton University Press.
Kalyvas, S. (1996) The Rise of Christian Democracy in Europe. Ithaca: Cornell University Press.
Kersbergen, K. and Waarden, F. (2001) Shifts in Governance: Problems of Legitimacy and
Accountability. The Hague: Nederlandse Organisatie voor Wetenschappelijk Onderzoek.
Kitschelt, H., Lange, P., Marks, G. and Stephens, J.D. (eds) (1999) Continuity and Change in
Contemporary Capitalism. Cambridge: Cambridge University Press.
Koenig, M. (2003) Staatsbürgerschaft und religiöse Pluralität in post-nationalen Konstellationen.
Marburg: University of Marburg, unpublished PhD thesis.
Koenig, M. (2004)  Öffentliche Konflikte um die Inkorporation muslimischer Minderheiten in
Westeuropa: analytische und komparative Perspektiven , Journal für Konflikt- und Gewalt-
forschung, 6(6): 85 100.
Koenig, M. (2005)  Incorporating Muslim migrants in Western nation states , Journal for
International Migration and Integration, 6(2): 219 34.
Laycock, D. (1997)  The underlying unity of separation and neutrality , Emory Law Journal, 46(1):
43 74.
Lijphart, A. (1984) Democracies. Berkeley: University of California Press.
Martin, D. (1978) A General Theory of Secularization. New York: Harper & Row.
886 V. Bader
Massignon, B. (2003)  Regulation of religious diversity by the institutions of the EU . Online at
(http://hirr.hartsem.edu/sociology_online_articles_massignon.html).
Mazower, M. (2005) Salonica. London: Harper.
Minkenberg, M. (2003a)  Staat und Kirche in westlichen Demokratien , Politik und Religion, 33:
115 38.
Minkenberg, M. (2003b)  Religion and policy effects , in Madeley, J. and Enyedi, Z. (eds) The Policy
Impact of Church State Relations: Family Policy and Abortion in Britain, France and Germany.
London: Frank Cass, 195 217.
Minkenberg, M. (2004)  Religious Effects on the Shaping of Immigration Policy in Western
Democracies . Uppsala: Paper given at the ECPR 32nd Joint Session, 13 18 April.
Monsma, S. and Soper, C. (1997) The Challenge of Pluralism, Church and State in Five Democracies.
Lanham and New York: Rowman & Littlefield.
Pierre, J. and Peters, B.G. (2000) Governance, Politics and the State. Houndsmill: Macmillan.
Pierson, P. (2000)  Three worlds of welfare state research , Comparative Political Studies, 33(6):
791 821.
Rath, J., Penninx, R., Groenendijk, K. and Meyer, A. (1996) Nederland en zijn Islam. Amsterdam:
Het Spinhuis.
Schmitter, P. (2000) How to Democratize the European Union? Oxford: Rowman & Littlefield.
Shahid, W.A.R. and van Koningsveld, P. (eds) (2002) Religious Freedom and the Neutrality of the
State: The Position of Islam in the European Union. Leuven: Peeters.
Treib, O., Bähr, H. and Falkner, G. (2005) Modes of Governance. EUROGOV Paper No. N-05-02.
Whitley, R. (2000) Divergent Capitalisms. Oxford: Oxford University Press.