The concept of integration as an analytical tool and as a policy concept

This chapter expands on recent approaches to the concept of integration. It defines integration in an open non-normative fashion as “the process of becoming an accepted part of society”. For the study of integration processes and policies a heuristic model is presented encompassing three analytically distinct dimensions in which people may (or may not) become an accepted part of society: (i) the legal-political, (ii) the socio-economic, and (iii) the cultural-religious. Each dimension involves different parties (the immigrants themselves and the receiving society), different levels of analysis (the individual, collective, and institutional levels), and other relevant factors such as time and generations. For the study of integration policies, policy frames, concrete policy measures, and both the vertical and horizontal aspects of integration policymaking are considered in order to account for the complex, multi-layered, and often contradictory character of integration policies. Use of the heuristic device enables integration processes and policies to be systematically described. However, comparison is key when the aim is to explain differences (and similarities) in integration outcomes. The conclusion returns to the concepts of integration and integration policies and suggests avenues for further research.


Introduction
The term integration refers to the process of settlement, interaction with the host society, and social change that follows immigration. From the moment immigrants arrive in a host society, they must "secure a place" for themselves. Seeking a place for themselves is a very literal task: Migrants must fi nd a home, a job and income, schools for their children, and access to health facilities. They must fi nd a place in a social and cultural sense as well, as they have to establish cooperation and interaction with other individuals and groups, get to know and use institutions of the host society, and become recognized and accepted in their cultural specifi city. Yet, this is a two-way process. The host society does not remain unaffected. The size and composition of the population change, and new institutional arrangements come into existence to accommodate immigrants' political, social, and cultural needs.
The scientifi c study of the process of settlement of newcomers in a host society has a long history. Popularized by the Chicago School of urban sociology in the early twentieth century, it has been approached from different perspectives and using a variety of concepts. A fi rst area of variation has to do with the object of study. Whereas some researchers have focused primarily or solely on the newcom-ers and (changes in) their ideas and behaviour, others have concentrated instead on the receiving society and its reactions to newcomers. A second area of differentiation lies in the dimensions of the process of settlement that are considered. Whereas some researchers have examined the legal and political dimensions of becoming part of a host society (e.g., legal residence, citizenship, and voting rights), others have concentrated on the socio-economic dimension (e.g., immigrants' access to health care, education, housing, and the labour market) or on cultural-religious aspects. Finally, the level of analysis has varied from that of individual newcomers and collective groups of newcomers and civil society to the institutional level, with questions being asked such as whether immigrant collectives have established their own institutions in the new society and, conversely, to what extent and how have institutions of the receiving society reacted to newcomers. While concepts such as adaptation, acculturation, and assimilation have tended to be focused on the cultural dimension of immigrants' settlement, others, such as accommodation, incorporation, and inclusion/exclusion, have shifted the focus to the host society and concentrated on the legal-political and socio-economic dimensions.
All of these approaches and concepts are highly contested within the academic literature. As for any term that stems from policy documents and debates, their definitions and the related discussions have been highly normative in nature. In relation to the concept of integration, the major point of criticism is the fact that it continues to assume-as did the old conception of assimilation-that immigrants must conform to the norms and values of the dominant majority in order to be accepted. This assumption elevates a particular cultural model, in the USA that of middle-class, white Protestants of British ancestry, and in many European countries that of a collectively claimed national language, culture, and tradition; a model that expresses the normative standard towards which immigrants should aspire and by which their deservingness of membership should continuously be assessed.
Integration is presented not only as a must but also as a straight-line process. Again, informed by policy discourses and policy goals, many studies of immigrant integration assume a more or less linear path along which the minority group is supposed to change almost completely while the majority culture is thought to remain the same. Nonetheless, as Lindo ( 2005 , 11) observes, taking integration as a selfevident and inescapable process ignores that the 'complex interplay of culturation, identifi cation, social status and concrete interaction patterns of individuals may produce many different "outcomes", much more varied in fact than a more or less linear shift from "immigrant" to "host" ways of doing things'.
Finally, the mainstream into which immigrants are expected and said to merge is seldom clearly defi ned (Favell 2003 ;Waldinger 2003 ). Some scholars argue that the concept of integration continues to adhere to an essentially functionalist vision of society in which immigrant success is still charted against a set of taken-for-granted mainstream norms bounded by the notion of a host society as a wholly self-contained unit of social processes (Gibney and Hansen 2005 ). Similarly, Joppke and Morawska ( 2003 , 3) observe that this concept 'assumes a society composed of domestic individuals and groups (as the antipode to "immigrants") which are integrated normatively by a consensus and organizationally by a state'. More recently, in the Dutch context, Schinkel ( 2010 ) coincides to note that the very notion of society is problematic, as it implies the existence of a more or less homogeneous and cohesive social environment in which only certain types of people-namely migrants-need to integrate.
Despite being a contested concept, integration continues to be central in many studies and debates on the settlement of newcomers in host societies. In Europe, several authors have attempted to strip the concept of its normative character and build a more open and analytical defi nition (Hoffmann-Novotny 1973 ;Esser 1980 ;Heckmann 1981Heckmann , 2015Penninx 1989Penninx , 2005Bommes 2012 ). Esser ( 2004 , 46) defi nes integration as 'the inclusion [of individual actors] in already existing social systems'. For Heckmann ( 2006 , 18), integration is 'a generations lasting process of inclusion and acceptance of migrants in the core institutions, relations and statuses of the receiving society'. According to Bommes ( 2012 , 113), 'the problem of migrant assimilation refers to no more (and no less) than the conditions under which they succeed or fail to fulfi l the conditions of participation in social systems'. In order to work or to gain access to goods, education, rights, and social welfare, Bommes (ibid.) argues, every individual must have some knowledge of what it means to work or how to behave as a patient, a client, a pupil, a student, or an applicant. From this perspective, there is no alternative to integration.
Interestingly, all of these approaches have in common the assumption that actors (immigrants in this case) are partially engaged in multiple autonomous and interdependent fi elds or systems. This implies a shift away from a holistic approach that conceptualizes integration into a taken-for-granted reference population-the "core culture" or national society as a whole-towards a disaggregated approach that considers not only multiple reference populations but also distinct processes occurring in different domains (Brubaker 2001 , 542-544). For instance, Esser ( 2001 , 16) refers to four dimensions: culturation (similar to socialization), placement (position in society), interaction (social relations and networks), and identifi cation (belonging). Similarly, Heckmann and Schnapper ( 2003 ) distinguish between structural integration, cultural integration (or acculturation), interactive integration, and identifi cational integration. From this perspective, integration dynamics and tempos are viewed as different for each dimension, and processes of structural marginalization and inequality become key.
In line with these more recent approaches to the concept of integration, this chapter aims to set up an analytical framework for the study of integration processes and policies. For this purpose, we focus in the fi rst part on the concept of integration, introducing an open non-normative analytical defi nition and identifying the main dimensions, parties involved, levels of analysis, and other relevant factors such as time and generations. In the second part, we defi ne integration policies and propose a distinction between policy frames and concrete policy measures as well as a shift from government to governance in order to account for the complex, multi-layered, and often contradictory character of integration policies. The conclusion returns to the concepts of integration and integration policies and suggests lines for further research.

A Defi nition of the Concept
We defi ne integration as "the process of becoming an accepted part of society" . This elementary defi nition is intentionally open in two regards. First, it emphasizes the process character of integration rather than defi ning an end situation. Second, in contrast to the normative models developed by political theorists, it does not specify beforehand the degree of or even the particular requirements for acceptance by the receiving society. This makes the defi nition highly useful for empirical study of these processes. Measuring the degree of becoming an accepted part of society will allow us to capture the diversity of (stages of) the process. We do need to specify within this basic defi nition what should be measured; that is, what are the indicators of integration and where might we fi nd them.

Three Dimensions
The basic defi nition of integration encompasses three analytically distinct dimensions in which people may (or may not) become an accepted part of society: (i) the legal-political, (ii) the socio-economic, and (iii) the cultural-religious. As pointed out by Entzinger ( 2000 ), these dimensions correspond to the three main factors that interplay with immigration and integration processes: the state, the market, and the nation. Focusing on these dimensions instead of the ones mentioned earlier (e.g., culturation, placement, interaction, and identifi cation) allows us to shift the focal point from immigrants to their relationship with a host society. The question is not only what immigrants do, with whom do they interact, and how do they identify themselves, but as much whether they are accepted and how they are positioned in each of our three dimensions.
The legal-political dimension refers to residence and political rights and statuses. The basic question here is whether and to what extent are immigrants regarded as fully-fl edged members of the political community. The position of an immigrant or the "degree of integration" has two extreme poles. One of these is the position of the irregular immigrant who is not part of the host society in the legal-political sense, though perhaps being integrated in the other two dimensions. The other is the position of the immigrant who is (or has become) a national citizen. In between there is enormous variety, which has increased in recent decades as a consequence of attempts of European states to "regulate" international migration and the new statuses and rights stemming from the EU migration regime (among others, EU nationals versus third-country nationals or "TCNs").
The socio-economic dimension refers to the social and economic position of residents, irrespective of their national citizenship. Within this dimension, the position of immigrants can be analysed by looking at their access to and participation in domains that are crucial for any resident. Do immigrants have equal access to institutional facilities for fi nding work, housing, education, and health care? Do they use these facilities? What is the outcome of immigrants' participation compared to that of natives with the same or comparable qualifi cations? Since needs and aspirations in these domains are relatively universal (basic needs are largely independent of cultural factors), access to and participation of immigrants and natives in these areas can be measured comparatively. The outcomes, particularly when they are unequal, provide useful inputs for policies.
The cultural-religious dimension pertains to the domain of perceptions and practices of immigrants and the receiving society as well as their reciprocal reactions to difference and diversity. If newcomers see themselves as different and are perceived by the receiving society as culturally or religiously different, they may aspire to acquire a recognized place in these respects. For their part, the receiving society may or may not accept cultural or religious diversity. Here again we fi nd two extremes. At one extreme, new diversity may be rejected and immigrants required to adapt and assimilate into mono-cultural and mono-religious societies. At the other extreme, ethnic identities, cultures, and worldviews may be accepted on an equal level in pluralistic societal systems. Between these two extremes again are many in-between positions, such as accepting certain forms of diversity in the private realm but not, or only partly, in the public realm.
This third dimension, and the specifi c positions of immigrants and immigrant groups, is more diffi cult to measure, basically for two reasons. Firstly, it is less about objective differences and diversity (ethnic, cultural, and religious) than about perceptions and reciprocal normative evaluations of what is defi ned as different and the consequences of such categorizations. Categorizations may become stereotypes, prejudices, and ultimately part of immutable racist ideologies. Moreover, the basis of categorizations may change. For example, in the guest worker period (1960)(1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975), the fact that an increasing share of immigrant workers were Muslims was not seen as relevant. It was only from the 1990s forward that such migrants and their families were categorized as coming from Muslim countries. Secondly, categorizations and reciprocal perceptions manifest differently at different levels (i.e., at the individual, collective, and institutional levels), and their consequences may also differ. If contacts between individuals are coloured by prejudice, interactions may be uncomfortable but have a limited impact. Yet, at the institutional level, if employers base their recruitment of workers on stereotyped or prejudiced perceptions and procedures, the consequences for individual immigrants may be quite negative.
It is important to realize that these three dimensions are not fully independent of one another. The legal-political dimension may condition the socio-economic and the cultural-religious dimensions (represented by arrows in Fig. 2.1 ). From the perspective of individual immigrants, factors such as illegal residence, extended uncertainty about future residence rights (compounded in the case of asylum seekers by long-term dependence on charity or the state), and lack of access to local and/or national political systems and decision-making processes have negative implications for opportunities and participation in the socio-economic and political realms. From the perspective of the receiving society, exclusionary policies are an expression of a general perception of immigrants as outsiders, which inevitably adversely affects immigrants' integration. The cultural-religious dimension may similarly impact the socio-economic dimension (represented by another arrow in Fig. 2.1 ). For example, negative perceptions of certain immigrants may lead to prejudice and discrimination by individuals, organizations, or institutions in the receiving society, and this may reduce immigrants' opportunities-even if access is legally guaranteed-in domains such as housing, education, health care, and the labour market.

Two Parties
Having defi ned the dimensions of the process of integration of newcomers into an established society and how to measure them, the next question is who are the relevant parties involved? Firstly, there are the immigrants themselves , with their varying characteristics, efforts, and degrees of adaptation (the left part of Fig. 2.1 ). Secondly, we fi nd the receiving society , with its characteristics and reactions to newcomers (the right part of Fig. 2.1 ). It is the interaction between the two, however, that determines the direction and the temporal outcomes of the integration process. However, these two "partners" are fundamentally unequal in terms of power and resources. The receiving society, especially its institutional structure and reaction to newcomers, is far more decisive for the outcome of the process than the immigrants themselves are.

Three Levels and Indicators
Processes of immigrants' integration take place and can be measured at different levels. The fi rst level is that of individuals , both migrants and natives of the receiving society. For the fi rst dimension, immigrants' integration at the individual level can be measured in terms of their legal status and political participation. For the second dimension, we can look at their socio-economic integration and position in the "hard" domains of housing, work, education, and health. For the third dimension, we would measure their identifi cation with a specifi c cultural-religious group and with the receiving society, as well as their cultural and religious practices and how these are valued. In our conceptual defi nition of integration, we should also measure the attitudes and behaviour (or acceptance) of native individuals towards newcomers and the consequences of these.
The second level is that of organizations . There are the organizations of immigrants, which mobilize resources and ambitions of the group. These organizations may be strong or weak; they may orient themselves primarily towards (certain aspects of participation in) the receiving society or to specifi c cultural and religious needs of the group. They may become an accepted part of civil society-and a potential partner for integration policies-or isolate themselves or be excluded by the host society. There are also organizations of the receiving society. Their extent of openness to newcomers, their perceptions of and behaviour towards individual immigrants, and their organizations might be of crucial importance for immigrants' integration. Research has shown, for example, that with the absence of governmental integration policy in Germany until 2002, nongovernmental organizations (NGOs), particularly trade unions and churches, played a crucial role in the integration processes of guest workers and their families (Penninx and Roosblad 2000 ).
The third level is that of institutions , understood as standardized, structured, and common ways of acting in a socio-cultural setting. Two kinds of institutions are of particular relevance. The fi rst are the general public institutions of the receiving society in the three dimensions: institutional arrangements of the political system; institutional arrangements in the labour market, housing, education, and public health; and institutional arrangements for cultural and religious diversity. Laws, regulations, and executive organizations, but also unwritten rules and practices, are part of these institutions. Though general institutions are supposed to serve all citizens in an equal manner, they may impede access or equitable outcomes for immigrants. They may exclude immigrants formally, either completely-as does the political system of most countries-or partially-as when social security and welfare systems offer only limited services to non-citizens. Yet, even if access for all residents is guaranteed by law, institutions may hamper access or equitable outcomes by virtue of historically-and culturally-determined ways of operating, for instance, by failing to take into account immigrants' history, their cultural and religious backgrounds, or their language abilities. Thus, adequate functioning of general public institutions-and their potential to adapt to growing diversity-is paramount. At this level, integration and exclusion are "mirror concepts" (see Penninx 2001 ).
The second type of institution that is particularly relevant for immigrants' integration is institutions specifi cally "of and for" immigrant groups, such as certain religious or cultural ones. Unlike general institutions, the value and validity of any group-specifi c institution is confi ned to those who voluntarily choose and adhere to them. Although their place is primarily in the private sphere, group-specifi c institutions may also manifest themselves as civil society actors in the public realm, as shown by the history of churches, trade unions, and cultural, leisure, and professional institutions in European cities and states. Some migrant-specifi c institutions may become an accepted part of society, equivalent to institutions of native groups. Others, however, might either isolate themselves or remain unrecognized or even excluded.
Different mechanisms operate at the individual, organizational, and institutional levels, but the outcomes at all of these levels are clearly interrelated. Institutional arrangements largely determine organizations' opportunities and scope for action, and they may exert signifi cant infl uence on how immigrant organizations develop and orient themselves. Institutions and organizations, in turn, together create the structure of opportunities and limitations for individuals. Conversely, individuals may mobilize to change the landscape of organizations and may even contribute to signifi cant changes in general institutional arrangements. In view of the uneven distribution of power and resources noted above, such examples are scarce but they are not nonexistent.

Time and Generations
The heuristic model developed and explained above may be used as a tool to describe and analyse the position of individual immigrants and groups of immigrants at a certain point in time. But an important element in the logic of integration processes is the time factor. Integration of newcomers is a long-term process by its very nature. This immediately becomes apparent if we look through the lens of newcomers. At the individual level, adult immigrants may adapt cognitively and adjust their behaviour when they learn how things are done, by whom, and so on. This part is relatively easy and pays off quickly. However, their adaptation in the aesthetic (relating to the fi ve senses) and normative realms takes more time. Feelings, likes, dislikes, and perceptions of good and evil remain rather persistent over a lifetime. Though this may be a general pattern for all human beings, it becomes especially manifest in those who have changed their environment through migration (for an overview of these aspects of the adaptation process see Van Amersfoort 1982 , 35 ff.).
The situation of the descendants of immigrants generally differs in this respect. Although they do become familiarized with the immigrant community and possibly its pre-migration background through their primary relations in family and immigrant community networks, they simultaneously become thoroughly acquainted with the culture and language of the society of settlement, not only through informal neighbourhood contacts starting in early childhood but especially through their participation in mainstream institutions, particularly the education system. If such a double process of socialization takes place under favourable conditions (in which policies can play an important role), these second-generation young people develop a way of life and lifestyle that integrates the roles, identities, and loyalties of these different worlds and situations. Because the ways of doing this are manifold, more and more differentiation develops within the original immigrant group. At the group level, this means that the litmus test for integration as an end result (being an accepted part of society)-and hence for the success or failure of policies in this fi eld-lies in the situation of the second generation in the host society.
In principle we can grasp the time factor by carrying out and comparing descriptive analyses of individuals and groups of immigrants at different points in time. In doing this, we should be cognizant of fi ndings of previous historical comparative analyses. First, research indicates that integration processes are neither linear nor unidirectional. Although we have indicated before that the situation of migrants (fi rst generation) differs signifi cantly from that of their children and grandchildren, this does not imply that integration is the inevitable eventual outcome. On the contrary, the literature shows that setbacks may occur. Second, we should keep in mind that integration may progress at different paces in the three dimensions and even within a single dimension-for example, labour market integration may take longer than integration in the health care system. Third, we should not forget the receiving society, which changes with immigration and has to adapt its institutions to immigrants' needs. For societies without a recent history of immigration or diversity, the process may be more demanding and therefore require more time than in immigration societies.

A Defi nition of the Concept
The study of policies is fundamentally different from the study of integration processes. The essence of policies is the intention to guide and steer processes in society, in our case, integration processes of immigrants. Explicit integration policies are part of a normative political process in which the issue of integration is formulated as a problem, the problem is given a normative framing, and concrete policy measures are designed and implemented to achieve a desired outcome. Other generic policies not specifi cally targeting immigrants (such as the education and health care systems, housing, the labour market, and the public regulation of religion) may exert strong infl uence (positive or negative) on integration processes of immigrants. Therefore, a systematic analysis of integration policies should go beyond integration policies in the strict sense.

Frames
When studying integration policies, the fi rst question to be analysed is how different political and social actors perceive immigrant integration in terms of policy frames and policy shifts . A frame is a reconstruction of the problem defi nition of a policy issue, including the underlying assumptions of the problem's causes and possible remedies for it. This means looking at how the problem is actually defi ned and explained and at what is thought could and should be done about it. The problem defi nition takes into consideration how immigration is perceived: Is it seen as a problem or as opportunity? Who has the moral or legal right to be or become an immigrant? Who are the wanted immigrants, and who are the unwanted? For those immigrants already present in the host society, a basic question is whether they are seen as "foreigners", as "temporary guests", or as permanent members of society for whom the state accepts the same responsibilities as for native citizens, guaranteeing the same rights and providing the same facilities.
Once the problem has been defi ned, the next step is considering what should be done . In some cases, a state or a city may choose to ignore immigrants' presence and therefore avoid any special responsibility for them. This choice for a non-policy response should be understood as a policy in itself (see Hammar 1985 , 277-278;Alexander 2007 , 37 ff). In other cases, new policies may be formulated to cater for certain immigrants' needs but under specifi c conditions due to the alleged temporary nature of their stay. Under this guest worker approach immigrants' otherness may be "tolerated" and even encouraged though their residence rights may be curtailed in the long run. Finally, if immigrants are perceived as permanent residents, inclusion is the main response. This takes different forms, however. Coinciding with the model on integration policies proposed by Entzinger ( 2000 ), integration policies may differ signifi cantly with regard to the three dimensions of immigrants' integration identifi ed earlier; that is, the legal-political dimension, the socio-economic dimension, and the cultural-religious dimension.
In terms of the fi rst dimension, legal recognition and political participation, policies may recognize immigrants as permanent foreign residents (the so-called "denizens"), thus incorporating them socially but limiting their political rights, or immigrants may be accepted as full citizens, thus removing all barriers for and even promoting naturalization. In terms of equality, the socio-economic dimension, specifi c policy measures may be devised catering for immigrants' interests and needs, or policies may merely address the common interests of citizens in general. Finally, in terms of diversity, the cultural-religious dimension, policies may be designed under two very different premises. The fi rst is that integration demands the adaptation and learning of immigrants but also signifi cant changes in access to and the working of institutional structures of the host society. The second is that societal rules and structures, including underlying norms and values, should be taken as a given and immigrants should (voluntarily or even as a mandatory task) adapt to them.
Finally, the third question to be addressed is for whom are integration policies meant . Migrant integration policies that designate specifi c groups of immigrants as target groups are different from policies that focus on all immigrants and are even more distinct from policies targeting all individuals regardless of their origin or targeting natives, established civil society, and the general institutions of society. In practice, these different approaches result in very different policies, again with regard to the three dimensions of integration: Political rights can be granted to immigrants as individuals, for instance, by granting voting rights, or as groups, which often means the creation of representative bodies. Policies may seek to promote equal opportunities for all citizens, meaning equal access to housing, education, health care, and the labour market, or they may seek to promote an equal share in access to these goods and services. Finally, cultural diversity can be promoted as an individual or group right, the latter often implying state support to immigrants' own organizations and institutions.
Frames cannot always be analysed directly; they often have to be reconstructed from policy documents and political discourse. When a policy is defi ned, it generally includes an explicit formulation of the perceived problem and the desired outcome of the specifi c efforts encompassed by the policy. Thus, politically debated statements in and about policy documents contain the essential elements of policy frames. The most important elements to be studied and compared are the general assumptions and orientations about the causes of the problem and remedies as well as basic concepts used (or explicitly rejected); the general aims of policies and dimensions of integration addressed; and the defi nition of the main target groups.

Policy Measures
Policy documents may be closer to policy discourse than to policy practice. In this regard, it is fundamental to complement any study of policy frames with a concrete and detailed analysis of actual policy measures. This means looking at the programmes in place and again identifying in which of the three dimensions of integration they are to be categorized, what their main goals are, and who they target. As said before, the study of integration policies cannot in general be limited to analysis of explicit integration policy measures. Programmes addressing the population as a whole or specifi c socio-economic groups within it, regardless of whether they are of immigrant origin, as well as general institutional arrangements in areas such as education, health care, housing, and the labour market, may be as fundamental (or even more) in fostering (or not) the integration of immigrants. Neither should we overlook how these policy measures are implemented in practice or to what extent and how street-level bureaucrats, practitioners, and professionals adapt them to their own goals and possibly limited resources.
In this regard, the study of policy measures entails a triple diffi culty: (i) we must go beyond integration policy measures in the strict sense, which greatly expands the fi eld of study; (ii) policy measures are seldom described in offi cial documents and therefore are diffi cult to trace; and (iii) programmes are often constituted of a set of unwritten norms and practices which may vary across time and space. A way to overcome these diffi culties is by conducting extensive fi eldwork and, especially, interviews with the main actors involved: policymakers at the different administrative levels, practitioners and professionals in the different social areas, NGOs, and immigrant organizations. When focusing on policy measures, it is also key to examine the budgets allocated in each programme in order to get a concrete picture of what actually is being done. Interestingly, policy frames and policy measures may differ signifi cantly in their goals, the dimensions of integration addressed, target groups, actors involved, and resources available.

Governance
Once we have identifi ed the main policy frames and policy measures, the next question is how integration policies are organized and implemented. Regarding organization, two aspects are relevant. The fi rst is whether the implementation of policies by civil servants and other actors is directly steered and controlled by politics or whether there is a relatively large gap between politics and policy. In highly politicized contexts, what politicians say and what actually is being done may differ signifi cantly. The second aspect of concern is the location of the initiating and coordinating force for migrant integration within the governmental administration: Is it centrally located and coordinated by a specifi c ministry or department (i.e., home affairs, social affairs, or employment)? Or is it decentrally organized across all of the areas relevant to integration policies. Such questions also apply to regional and local policies (Caponio and Borkert 2010 ).
If we want to examine not only how policies are organized but also how they are formulated and implemented, we should shift the focus from government to governance. This means taking into account a wider range of actors, including other administrative levels such as regional and local governments; other institutions, agencies, and practitioners within the state apparatus; and other relevant actors, such as politicians, NGOs, and private institutions. The vertical dimension of integration policymaking, that is, the relationship between the national, regional, and local levels, is of particular importance, as both municipalities and the European Union (EU) level have become increasingly involved in the making of immigrant policies. This multiplicity of levels should be analysed in detail so as to understand how new tensions have come to the fore but also how new alliances and forms of cooperation (e.g., between the local and the EU level) have developed. Various key questions can be asked: Who is in charge of integration policies? How are the different levels coordinated? Do they respond to different political and social imperatives? Do they complement or contradict one another?
Also to be considered is the horizontal dimension of integration policymaking, meaning whether and how integration policies are implemented by the full range of relevant actors, from private institutions to NGOs, immigrant organizations, and professionals. The central question here is who is supposed to be a relevant actor in policies. With respect to immigrants, are individual immigrants seen as primary actors? Are their organizations and other collective and institutional resources regarded as relevant? Looking at the receiving society, what main actors are involved, again at the individual, organizational, and institutional levels? Research on Southern Europe has shown that when governmental integration policies are absent, civil society actors (such as trade unions, NGOs, charities, and civil movement associations) may become key in providing various services and offering political support for immigrants' rights claims (Campomori 2005 ;Zincone 1998 ). At the same time, as noted by Caponio ( 2005 ), such mobilization may produce a "crowding out" effect wherein native associations mobilizing on behalf of immigrants actually become the main recipients of municipal funding and partners in policymaking. Immigrants may thus be prevented from forming their own organizations.

Politics and Time
In democratic societies, policies are part of a political system in which the majority decides. This brings an inherent danger of either a virtual absence of explicit integration policies and an avoidance of issues related to immigrants or one-sided patronizing policies refl ecting mainly majority interests and disregarding the needs and voices of immigrants. Whereas in some European countries policymakers have been able to craft policies "behind closed doors" to extend political and social rights to migrants (Guiraudon 2000 ), in others anti-immigrant political parties have succeeded in vetoing liberal reforms and urging their governments to adopt more restrictive immigration and integration policies. An extreme case is Switzerland, where referendums can even overrule the supreme court and possibly mandate reform of the constitution, thus undermining the main tools that protect religious and ethnic minorities of immigrant origins against discrimination (D'Amato 2012 ).
As integration policies are adopted and implemented in practice, another aspect of the logic of policymaking emerges. Although integration processes are long term in nature-they take at least a generation-the political process in democratic societies requires that policies bear fruit within much shorter timeframes: the spaces between elections. Such a policymaking context may lead politicians to make unrealistic promises that cannot be fulfi lled in such a short period. This "democratic impatience" in turn often produces disappointment and backlash effects (Vermeulen and Penninx 1994 ). The debate on the alleged failure of integration policies-and of immigrants to integrate-that has been taking place in the Netherlands since 2000 is a good example (Prins and Saharso 2010 ). Even more diffi cult than democratic impatience are situations in which anti-immigrant sentiments are translated into political movements, leading to strong politicization of the topics of immigration and integration.

Integration Processes
If the immigrant integration process is propelled by interactions between immigrants and the receiving society at the different levels, within the three dimensions and taking into account the time factor, the best way to explain diversity in outcomes is through comparative empirical studies. There are two main types of comparisons, each measuring different elements of our heuristic model. In the fi rst category are studies that compare the integration processes of different immigrant groups in the same institutional and policy context of a nation or a city. Such studies reveal that different immigrant groups may follow different paths of integration. For example, in the Netherlands, Vermeulen and Penninx ( 2000 ) show that Moluccan, Surinamese, Antillean, Southern European, Turkish, and Moroccan immigrants differ in the speed of their integration and in the pathways they follow. Whereas some groups (e.g., the Chinese and Portuguese) have been quick to use the education system as a route to social mobility, other groups (such as the Turks) were more strongly involved in entrepreneurship. The consequence of such comparison is that the factors found to explain differences lie primarily in the characteristics of the various immigrant groups (thus, the left side of our heuristic model), simply because the national or local context into which immigrants are being integrated is the same.
A second category of comparative studies examines the integration of the same immigrant groups in different national or local immigration contexts. Koopmans ( 2010 ), for example, investigates the effects of integration policies and welfare state regimes on the socio-economic integration of immigrants in eight European countries. The comparison leads this author to conclude that multicultural policies, when combined with a generous welfare state, produce low levels of labour market participation, high levels of segregation, and a strong overrepresentation of immigrants among those convicted for criminal behaviour. Another study of this kind is The Integration of the European Second Generation (TIES) project, a comparative analysis of the position of the children of Turkish, Moroccan, and Yugoslavian immigrants in 15 cities in eight European countries (Crul et al. 2012 ). That research asked how we might explain the higher educational attainment of second-generation Turks in Sweden and France compared to that in Germany and Austria, and why attainments are different when it comes to access to and integration into the labour market. One of its conclusions is that the contextual conditions created by institutions (e.g., schooling arrangements and labour market, citizenship, and welfare policies) are paramount to explain differences in educational and labour outcomes. Comparisons examining the same groups in different contexts tend to fi nd the main explanatory factors residing in the receiving society and mostly at the institutional level (the right side of Fig. 2.1 ).

Integration Policies
Comparative studies are also a tool to understand what conditions account for the emergence of different integration policy models as well as the factors that explain recent trends of convergence in both policies and practices. To explain such differences, various typologies have been developed. One of the most cited is the study by Brubaker ( 1992 ) of citizenship policies in France and Germany arguing that the different nation-building histories of France and Germany have led to distinctive conceptions of citizenship. Focusing on the degree of accommodation or acceptance of minority group cultures, another highly cited categorization is that by Castles ( 1995 ), which distinguishes between differential exclusion, assimilation, and pluralism. Though these typologies are based on rich historical accounts of integration policy development in different European countries, their relevance has been increasingly questioned. The multitude of national models of integration policies in existence has been criticised for overlooking the importance of the transnational and local levels as well as for minimizing internal incoherencies and changes over time.
During the past decade, comparative studies have rendered the analysis and explanation of integration policies signifi cantly more complex by taking into account the supranational (particularly European), regional, and local levels; by focusing on particular policy domains; and by examining the impact of a set of compound factors such as politics and the party system, the constitutional courts and judiciary power, welfare state regimes, and the role of civil society, the media, experts, and civil servants. At the local level, Alexander ( 2003Alexander ( , 2007 was one of the fi rst scholars to look at the city as the central unit of comparison, building a theoretical model to explain local policy reactions to migrant settlement over time and across a wide spectrum of cities and policy domains. Based on the concept of "hoststranger" relations, he distinguished four types or phases of local policies: a nonpolicy, a guest worker policy, an assimilationist policy, and a pluralist policy. Though Alexander's typology has been criticised as teleological, as well as for paying insuffi cient attention to policymaking and implementation (Caponio 2010 ), his comparative model is still a key reference in the literature on local integration policies.
An early example of comparative studies on particular policy domains is the comparison of the institutionalization of Islam in the Netherlands, Belgium, and the UK in the post-war period by Rath et al. ( 2001 ). Among the questions posed by that research were to what extent are Muslims being given the opportunity to set up their institutions according to their own agenda and how are public manifestations of Islam regulated, like mosque building and Islamic religious education in schools? These authors fi nd signifi cant variation in the institutionalization of Islam in timing, content, and direction in the three countries considered and between different cities. Looking at policy implementation and bureaucratic practices, comparative studies show important trends towards convergence. Particularly with regard to access to services, civil servants, NGOs, and professionals seem to respond similarly to similar everyday pressures, regardless of very distinct policy contexts. For instance, in a study of how Amsterdam and Berlin policymakers and policy practitioners deal with youth unemployment among immigrant groups, Vermeulen and Stotijn ( 2010 ) point to important similarities in terms of targeting relevant groups regardless of whether local governments pursued general or targeted policies.

Conclusion
We opened this chapter with a paradox: While many scholars reject the concept of integration arguing that it is highly normative and teleological in nature, the concept of integration nonetheless continues to be central in many studies and academic debates. How can we solve this contradiction? How can we study the process of settlement of newcomers in host societies and policies aiming to foster this process without falling into the pitfalls of the old assimilation/integration approach? With these questions in mind, this chapter presented a heuristic model for the nonnormative, analytical study of both integration processes and policies. First, we proposed a disaggregated approach to the concept of integration, distinguishing three dimensions (the legal-political, the socio-economic, and the cultural-religious), two parties (the immigrants and the receiving society), and three levels (individuals, organizations, and institutions). Second, for the study of integration policies, we suggested taking into account policy frames, concrete policy measures, and both the vertical and horizontal aspects of integration policymaking.
While the use of this heuristic device enables a systematic and analytic description of integration processes and policies, comparison is key when aiming to explain differences (and similarities) in outcomes. In the past decade, a number of research projects have compared integration processes by focusing either on different immigrant groups in the same national or local context or on the same immigrant group in different contexts. Integration policies have also been objects of comparison. While most early studies focused exclusively on the national level, more recent approaches have taken into account the supranational and local levels, particular policy domains, and concrete implementation practices by street-level bureaucrats and practitioners. Though these researches have signifi cantly contributed to the understanding of integration processes and policies, there is still much to be done.
Looking at integration processes, new systematic comparative analyses might shed more light on how particular immigrant cultures and migratory histories on the one hand, and general public institutions and immigrant policies on the other, shape integration outcomes. If we look at integration policies, comparisons between different cities and regions in the same country and in different countries are key to enable us to understand not only local and regional policy responses but also the relationship between the national, regional, and local levels. Despite diffi culty in terms of fi eldwork, more research is also needed on policy implementation practices. Comparisons of these will enable us to elucidate and understand important differences between policies as written and policies as practised as well as to identify and explain trends of convergence in this regard. Finally, while comparative research on integration processes has been done in North America and Europe, most comparative research on integration policies has been limited to Europe. Going beyond these traditional geographies of comparative studies will be essential to understand how more signifi cant differences in terms of nationhood, welfare state, or public and immigrant policies can lead to different outcomes in terms of integration processes. Looking in from the geographic outside may help to defi nitively strip the concept of integration of its normative and, above all, Western-centric character.
Open Access This chapter is distributed under the terms of the Creative Commons Attribution Noncommercial License, which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.