The Multi-Level Governance of Intra EU Movement

Free movement management is a challenge that emerged in a multi-level context, with policies developed at one level having unclear implications for actors working at other levels, and with questions of authority and competencies remaining highly uncertain. Broader issues of national and local contexts, as well as traditional institutional practices, all lay the groundwork for the need of clearly articulated governance solutions. As is clear from previous chapters, the flows of CEE migrants, as well as the social consequences of their presence in urban regions that are often unequipped in both policy and administrative terms, has led to substantial challenges for actors at all levels of government, both public and private, when it comes to free movement management.


Mapping Free Movement Management: From the Multi-Level Governance Literature to Modes of Governance
On one level, it is unsurprising to consider the case of free movement management in urban regions against the backdrop of the broad multi-level governance approach. In general, the broad multi-level governance literature posits that there is a multiplicity of actors present from the EU, national, regional and local levels, who engage in what can be regarded as a process of negotiation, deliberation and implementation that is more or less continuous (Schmitter and Kim 2005). The relationship among actors at these four levels is one where there is substantive engagement of those involved, not merely presence, and where engagement allows for the possibility of genuine mutual influence, but also facilitates and mutual dependency as various policy-making activities become intertwined with one another (Stephenson 2013). After all, the challenge associated with developing and implementing the tools to successfully manage free movement can be thought of as occurring in a clear-multilevel context. In a stylized nutshell, the principle of free movement of people within the EU often creates substantial challenges for urban actors (both public and private) who find their efforts to "solve" the issue difficult by a lack of clear policies at both the national and local level for addressing the issue. Coupled to this, CEE migrant integration within EU member states is regularly characterized by ambiguous lines of authority as to which actors have (or even want) responsibility for formulating and implementing policies. As such, given the multi-level context of the challenges associated with free movement management, one might suspect that it is a prime candidate for a governance solution that is fittingly multi-level in response. Indeed, some authors within the multi-level governance literature, as well as actors in the public sphere -in particular the EU itself -have worked from an assumption that multi-level governance is, by default, a desirable outcome where the problem context can be framed as multi-level in nature. Such excessively positive portrayals of multi-level governance have been called into question by scholars over the past decades, with arguments being made that multi-level governance has the potential to compromise democracy (Guy Peters and Pierre 2004), in essence undermining the principles of traditional government through facilitating for the establishment of self-organizing and self-regulating networks apart from traditional governing institutions (Harlow & Rawlings 2007).
However, as a number of studies focusing on urban administrative responses to the challenges of immigration have shown, the complexity of a multi-level system entailing many policy processes might not result in effective multi-level coordination at all. Rather, a broad variety of governance responses are possible, with the possibility that actors at any one level could play the decisive role in formulating and coordinating policy and administrative measures, or even, that no patterns of coordination might emerge. (Bowen 2007;Joppke 2007;Ostaijen et al. 2016;Duyvendak and Scholten 2012;Scholten 2012).
Clearly then, a response of multi-level governance to the challenge of free movement management is one possible outcome, but it should not be assumed to be the only or most likely governance result, solely because of the multi-level context in which actors find themselves embedded. What then are the other possible alternatives that might emerge?
In this chapter, we employ a typology of governance approaches that capture the full range of possible responses from the various public and private actors who are faced with developing responses to the challenges of free movement management. This typology of governance responses is comprised of the following four modes: horizontal governance, top-down governance and multiple-level governance and multi-level governance. We envision an overall governance setting comprised of four possible levels: the EU, national, regional and local. The actors inhabiting these levels may be either public (governments or administrative bodies) or those from the broader polity (parties, NGOs, interest groups, social partners, media, knowledge actors, etc.). This typology is more fully developed in the introductory chapter of this book, but we provide readers who might make use of this chapter in isolation from the others with a sufficiently detailed overview.
Horizontal governance is characterized by an engagement and participation among actors at only one level seeking to resolve free movement management challenges. While these actors may need to work within the confines of formal policy established at another governance level (most notably in the case of horizontal governance at either the local or regional level, who often must coordinate their activities against the backdrop of certain national policy mandates), formal and informal engagement is solely limited to one level of actors.
Top-down governance is characterized by a hierarchical relationship in which the highest level involved steers the overall governance response for actors from all other involved layers. In order for there to be a top-down governance response to the challenge of free movement management, there must be at least two levels involved, though these levels do not necessarily need to be adjacent. For example, in a setting where only actors at the national and local level are involved in developing free movement management measures, it can only be regarded as top-down governance if the national level has the ability impose solutions on local actors who are also involved in the deliberation process.
By contrast, bottom-up governance is characterized by a hierarchical relationship in which the lowest level involved steers the overall governance response for actors from all other involved layers. In order for there to be a bottom-up governance response to the challenge of free movement management, there must be at least two levels involved, though these levels do not necessarily need to be adjacent. For example, we witness bottom-up governance if the local level has the ability to impose solutions on regional or national -level actors who are also involved in the deliberation process.
However, it is imperative to stress that central to either top-down or bottom-up governance is the presence of a mutual collaborate relationship among actors at different levels, which is a necessary condition of governance. Lacking such a collaborative relationship, what may appear at first glance to be either top-down or bottom-up modes of governance is, in fact, a form of policy learning, in which actors at various levels simply adopt the governance solution developed horizontally at some other level, without any substantive engagement among these levels in constructing the specific governance response. We return to this point in the conclusion, highlighting the manner in which the repeated presence of horizontal governance at the local level across our cases ought not to be misinterpreted as evidence in support of bottom-up governance.
Multiple-level governance is characterized by a more or less equal relationship in terms of engagement and participation for actors at those levels involved in the process of responding to the challenges of free movement management. In order for there to be multiple-level governance, there must be at least two levels involved, though these levels do not necessarily need to be adjacent. However, as will be explained shortly, no more than three levels can be involved. For example, a setting where actors from national, regional and local levels are involved in developing free movement management measures can be regarded as multiple-level governance if there is a more or less equal relationship among actors from these different levels.
Lastly, there is multi-level governance in the traditional way that is understood in the literature: a situation where there is a more or less equal relationship in terms of engagements and participation among all four levels -including the EU level -as they seek to formulate responses to the challenges of free movement management.
As we turn to the major policy areas that have been the focus of free movement management responses in each of our four cases, we will see that multi-level governance, the one that has received the most attention in the literature, is that which is primarily lacking. Rather, the governance modes that are opted for in each of the major policy areas have exclusively been top-down, multiple, or horizontal in nature. It is only when we look beyond the key components addressed in each case responses that we see clear evidence of a multi-level governance response.
Following this, we intend to shift the theoretical discussion to matters of governance approaches more broadly, calling attention to the explanatory utility associated with network approaches (Curry 2015), as well as to horizontal, vertical and diagonal approaches (Torfing et al. 2012). We will stress the manner in which this combined literature highlights the importance of governance approaches as venues for interaction among actors at a variety of levels and within a variety of organizations -both state and civil society. The aim is to set the stage for a characterization of governance approaches that we regard as analytically straightforward for capturing the full range of possible governance relationships that can exist among and across levels: multi-level governance in the standard sense of the term within the literature: a largely equal relationship in terms of engagement and participation among actors at all four levels; top down governance, in which there is a hierarchical relationship steered by the highest involved level and directed downwards; a multiple level governance, in which there is an equal relationship in terms of engagement and participation among only those levels involved; and a horizontal governance approach, in which engagement and participation among actors at the same level can be observed (Zelano et al. 2016a, b).
The analysis that follows maps the policy measures taken in the different policy domains for which CEE migration have implications (socio-economic, sociocultural and legal-political), characterizing them in accordance with the typology presented above. In order to reconstruct urban governance approaches, we have first mapped the relevant stakeholders. This was done by each country team for the two selected urban regions per country. Following the identification of relevant actors, a reconstruction was made of (a) their practices and measures with regard to CEE migrants and (b) the rationale behind these practices and measures. This was achieved through interviews as well as both primary and secondary document analysis. As such, the methodological strategy has allowed us to capture both formal policy responses as found in texts and the governance approaches that can be witnessed in actual practice.

Austria: Mostly Multiple-Level Governance, but Also
Horizontal 1 The predominant area in terms of the Austrian case and specific areas focused on developing policy instruments for free movement management at the urban level is that of the labour market. While a variety of instruments are evident in this sphere, they are united by the fact that they by and large do not consist of measures consisting of multi-level governance type responses. In general, these responses can be characterized as instances of multiple-level governance, reflecting the degree to which there has been a strong presence of the national level in those cases where legislation has been central to the response, and also the degree to which national responses have been mirrored by related activity at the local level. The 2011 Anti-Wage Dumping Law, introduced by the Federal Ministry of Labour, Social affairs and Consumer Protection (BMASK) represents a multiplelevel approach response to the integration of CEE migrants. This legislation, stemmed from concerns over the wage gap between the Austrian labour market and that of new EU member countries. In 2015, the law was amended so that violations were dealt with as matters of administrative and not civil law, with the result being the emergence of new networks to manage and make sense of the law's requirements. Such networks took shape at the local level, for example in Vienna, where the Chamber of Labour serves as a source of information and a point of coordination for the various units in the city administration whose work with the law on a regular basis.
Multiple-level governance responses also tend to be characteristic for the variety of measures intended to deal with the phenomenon of CEE migrant de-qualification upon relocating to Austria and seeking opportunities on the labour market. Similar to the measures dealing with the prevention of wage-dumping, the efforts targeting migrant de-qualification see BMASK assuming a central role. One example of this are the four Contact Points for the Recognition and Assessment of Qualifications Obtained Abroad (AST), established in 2013. These contacts points serve as regional centres that offer guidance and counselling in a variety of languages, free of charge, for those seeking to have the foreign qualifications validated. Yet, what makes these regional contact points examples of a multiple-level governance response is not their location across various regions of Austria, but rather the actors who are charged with running the centres themselves. Rather than this being under the direct auspices of BMASK, these four contact points are managed by both NGOs and local administrative actors, resulting in an initiative that has seen state, civil society and local public actors collaborate to deliver services to CEE migrants requiring help with validation of foreign qualifications. A similar multiple-level governance response addressing CEE migrant de-qualification is the Network for the Recognition of Qualifications, established by the Austrian Integration Fund, which brings together social partners and professionals to discuss specific measures that could be introduced to improve the process of validating foreign qualifications, and also providing an information platform for cities and regions seeking relevant information about the validation process.
Taken jointly, these examples hammer home the prevalence of a clear pattern: in devising solutions to address the labour market challenges associated with free movement management, there exists primarily close working relationship between national and local level -initiatives developed at one level find mirrored counterparts at the others, and there is substantial interaction between the two levels in addressing how labour market challenges are to be resolved within urban areas.
By contrast, the other key area of free movement management in Austria -housing -is characterized exclusively by horizontal governance, taking place primarily at the local level. Yet, this difference from the area of labour market is not simply in terms of the mode of governance. Rather, the focus on housing at the local level in Austria is characterized by measures that are not policy-specific in nature, nor are they exclusively targeted at CEE migrants. For example, in the Austrian capital, the housing situation for migrants, especially those who might be characterized as economically vulnerable, is made challenging by the fact that social housing is only available to those who have been permanently resident in Vienna for two years and who are in possession of a registration certificate. Horizontal governance measures at the local level that seek to respond to this challenge more generally include the "Immigrant Fund" which can quickly provides low-threshold housing for migrants, assuming they possess the requisite security deposit, or the short-term housing measure KuWo, which provides the homeless ill with a place to recover following being discharged from the hospital. In both instances, the actors involved in these measures represent may of the key organizational players at the local level, including social partners, NGOs, and the city of Vienna itself. When it comes to measures that are EU-migrant specific, two in particular stand out: the Centre for Homeless EU Citizens (Zweite Gruft) and the night shelter for male EU citizens "VinziPort". While both represent horizontal-level responses to deal with the problem of housing for vulnerable EU migrants, the NGO Zweite Gruft has an even broader mandate, with activities not just cantered around housing, but also providing mediation for EU migrants who need assistance when engaging with various public authorities.
As such, the governance network that has taken shape in this other primary area of free movement management is circumscribed to actors at the local level. A rich illustration of programs and policy initiatives all serve to underscore that, while there is substantial policy activity, it does not involve engagement among actors from different levels. Housing is acted upon solely in local terms.

The Netherlands: Business as Usual: Local Level Horizontal and Multiple Level Governance Approaches
In the absence of any multilevel governance approaches in the Dutch case, what do we find? Primarily, free movement management has been met with horizontal and multiple level governance approaches. Housing issues, labour market considerations and question about civic registration were the three primary policy areas which triggered extensive governance responses by Dutch policymakers. Moreover, these were soon defined as 'CEE issues' in the sense that administrators and politicians regarded them as meriting special treatment compared to how housing, labour market and registration were normally dealt with. Governance actors have defined housing, labour market and registration as areas where CEE migration has created new challenges as parts of it tends to be temporary and circular. The approach taken by the officials was in many cases that of trying to minimize problems that Dutch citizens expressed about CEE immigrants. To some extent this particular angle was a result of the long established Dutch type of corporatism, which is built on active inclusion of all recognized and organized communities in society. As CEE migrants have not developed any strong communities the Dutch institutional logic contributed to excluding them. Other issue areas never seem to have evolved into CEE issues as such but have rather been funnelled by the administrative-political actors into already existing general policies. Examples of these were health care and education.
While the overall governance responses in the two Dutch case regions have been remarkably similar, there are still substantial variations inside each region. To some extent these differences have had to do with differences regarding different social consequences in different geographical areas.
A general observation is the fact that several policies and approaches began as municipality projects. In most instances, issues regarding CEE migration became evident at the local level, in the sense that immigrants literally encountered housing and employment problems in a certain city or village rather than in a government Ministry. Local level officials simply had to deal with issues arising at their doorstep. Later, local officials developed horizontal approaches by involving various local stakeholders -such as civil society, other Dutch municipalities and cities in neighbouring countries. In some cases regional and national level officials have stepped in as well.
A key aspect is whether policymakers have been able to fit CEE issues into existing local, regional or national level governance structures. Housing is normally dealt with at the local municipal level. Thus, a range of city policies ensued including a gradual development of horizontal cooperation between cities and other local level actors. However, since the Dutch Ministry of Internal Affairs and Social Affairs is responsible for overseeing the housing, labour market and registration areas some policies were developed at the national level as well. As a consequence, the governance responses have had elements of both local horizontal and multiple level features.
Housing remains an exceptionally pressing issue. Both the selected Dutch regions have been very active to combat various forms of housing problems such as abusive and non-compliant landlords. One particular concern has been to prevent two kinds of overcrowding: in flats and in neighbourhoods. A flurry of city initiatives have followed and in some cases elements of the local solutions have been picked up nationally and made into national law. However, these processes have rarely developed into mutually reciprocal systems of cooperation but there are exceptions, e.g. the 'Approach to attack slum landlords' (an agreement between municipalities and the Ministry of Internal Affairs). Thus, the overall governance approach in this regard has been local level horizontal with some elements of multiple level governance.
Dutch governance responses regarding labour market aspects have had a strong horizontal governance character. Rotterdam and the Hague have developed ties with each other and with other local level institutions such as by signing special agreements ('Covenants') to control fraud and mala fide employment arrangements.
To a lesser extent, an element of multiple level governance exists as well. Indeed, municipalities have collaborated to put labour migration on the national level agenda. The national government has increasingly been involved in issuing work permits and has introduced new legislation aimed at combatting wage dumping and at improving working conditions. Moreover, the national government has tried to coordinate actions against social dumping and other negative labour market aspects. While government ministers have been quite active, their efforts have mostly resulted in an element of coordinated agenda setting actions in a number of EU countries. Open letters have been written, op-ed articles have been published and conferences have been organized but little in terms of horizontal national level policies as such has been developed.
The general pattern of the Dutch case then, is that while local level responses may dominate, there are also clear instances of multiple-level strategies intended to assist in the management of free movement challenges. In the areas considered most salient for devising governance responses, networks and actions that are multi-level in nature are, by and large, lacking.

The Case of Turkey: Top Down
The governance approach in Turkey regarding CEE migration can best be summarized as top down. However, while Turkey is a unified state with a strongly state centred structure a multitude of laws and legislations contribute to inconsistent policies. Immigration policies are covered by jurisdictionally complex and fragmented legal orders. Due to this non established character, central authorities begin to develop their own policies, which sometimes clash with those of other central authorities. The result is an ad hoc, vague governance approach within an overall top down setting. Primarily immigration policies have been focusing on labour market and registration issues, while social security, health care access and housing issues have been more peripheral.
Within the overarching format, some differences exist between Edirne and Istanbul. Primarily these differences are down to differences regarding two aspects: the character of the migrants and the situation of the cities in a wider Turkish context. The bulk of Edirne's migrants are in fact well established CEE migrants (most of Turkish ethnic descent) of high socio-economic status. Istanbul on the other hand has become a major centre for Syrian refugees. Moreover, migration policies regarding Edirne are influenced by special border security considerations. As a consequence few specific CEE migrant policies have been developed in Edirne. Giant Istanbul attracts many immigrants. The various labour market regulations developed for immigrants do not differ between different migrant types. As with most policies policy coordination between different parts of the central authority structure has proved to be difficult although some data indicate efforts to strengthen and streamline central authority.
Some policy domains are absent or almost absent. For instance, very few policies have been developed regarding housing apart from some shelter provision in Edirne and some facilities regarding children in Istanbul. This absence reflects a traditional attention on labour market and registration aspects in relation to immigration. However, there appears to be a growing awareness among policymakers and other stakeholders that education, social security, healthcare and integration need more attention. To some extent the policy vacuum may disappear as a result of the heavy increase in Syrian migration. Since most of the existing policies so far have not differed between immigrant groups any development may well then affect CEE migrants as well.
Future developments will probably rely on which party or coalition of parties will hold the national majority. So far, the governance of migration has been clearly dependent on politics. The governance approaches have not been solidly institutionalised and have been subject to frequent changes. Moreover, although registration has been a primary domain on which policies have been developed this has not meant that a well-coordinated and systematic system has emerged. A bewildering array of rules, legislations and concept exist. Moreover, due to political and other tensions among policymakers even basic concepts such as 'asylum seeker', 'refugee' and 'migrant' have not become unequivocally defined even inside the top state level.
As such, given the nature of Turkish political institutions, it is perhaps the easiest case in which governance patters can be identified. Even the most cursory review of policies and networks suggests that there is an exceptionally close correspondence between, on the one hand, the institutional logic of the Turkish political system and the resulting top-down manner in which all aspects of managing migrant integration are constructed.

Sweden: Horizontal and Top-Down Predominate
Comparing the governance responses across the urban regions in Gothenburg and Stockholm, the group of destitute, often unregistered CEE citizens and beggars were at the centre of debates and governance responses. Homelessness of CEE citizens, most of them beggars from Bulgaria and Rumania, was one of the critical issues that required local governance responses. An institutional obstacle encountered by homeless EU citizens is their exclusion from the 'shelter guarantee' normally offered by Swedish municipalities. Since EU citizens are excluded from the shelter guarantee a local horizontal governance approach emerged to deal with the issue. The help and support for homeless EU/CEE citizens is offered through churches and charities, through formalized partnership with municipalities. The development of tighter and more frequent relations between the municipal level and the voluntary sector is a sign of new governance patterns set in motion by the novel social implications. Recent developments have also seen central authorities evicting homeless people who have set up temporary tent camps on private property. When faced with an increasing number of cases concerning illegal settlements, the Swedish Enforcement Agency ordered the regional units to contribute to a central, statistical record and initiated a project to harmonize local practices. Thus, while the governance approach regarding homeless CEE citizens is more or less local and horizontal there are some elements of top down activities as well.
Issues related to social security/help and registration of CEE EU citizens were also highly relevant for particular governance responses. These are two policy areas characterised by inconsistent practices, and bureaucratic uncertainty. Relevant documents and interviews show that Swedish municipalities seem to be at a loss about how to treat EU citizens who are living in Sweden and who are in need of social support. The municipal administrations find it difficult to decide whether an EU individual should be granted right of residence, which is in turn related to whether the individual can be granted social support on the same terms as Swedish citizens. A recurring theme has thus been for municipalities to call for clearer national guidelines regarding administrative procedure. In this area, we can observe development towards multiple governance approach. In an attempt to meet the demands of confused and uncoordinated local authorities, the Swedish Association of Local Authorities and Regions (SALAR) put together a handbook, including among other things, guidelines on how to assess applications for economic assistance by EU citizens. The National Board of Health and Welfare has also published a handbook, about how the national and the EU rules are to be interpreted by municipalities. In January 2015 the government appointed a national coordinator to improve the situation for the many EU migrants who are begging on the streets in Sweden. The role of the national coordinator is to work closely with, and gather information from, different national and international voluntary organisations and municipalities in order to improve the way they work together.
Similarly to the issue of social support, the registration procedure of EU citizens was subject to inconsistencies. In Sweden, EU citizens on short term contracts (or self-employed) can apply for a so-called Co-ordination Number (Samordningsnummer). In 2010 the Swedish Unemployment Insurance Board (a Government audit agency) highlighted inadequacies concerning the Jobseeking Agencies' EU citizen Co-ordination Number procedures. These citizens were unable to receive jobseeker-employer matching support from the Jobseeking Agencies on equal terms with Swedish citizens. However, it was not until spring 2013, when the issue was brought to the agenda by Swedish media that Swedish Jobseeking agencies promised to act in order to deal with the irregularities. To deal with this specific problem a national horizontal governance approach was established. The Swedish Tax Agency cooperates with other major national public agencies, e.g. the Employment Agency, and the Social Insurance Agency. Twice a year, and when needed, the Tax Agency and the Migration Board organise coordinating meetings, and keep in touch on a regular basis between the meetings. The Tax Agency also meets the National Board of Trade, the Social Insurance Agency, the Employment Agency and the Swedish Council for Higher Education about twice a year.
The blurred line between formal and informal employment create challenges in the receiving urban regions with informal work showing consequences such as wage related discrimination, exploitation, inferior working conditions and marginalisation. Here we can observe a governance mixture with some top down features and an element of national horizontal governance involving trade union and business sector cooperation. An issue addressed outside the reach of local public administrations, is the issue of posted workers from the CEE. Employers and trade unions disagree about the scope, and the implications posting brings to the Swedish labour market. Both have contact with public authorities at the national level. The four committees commissioned by the government to deal with various aspects of the posting of workers, are also an example of a top down approach. The presence of CEE workers has changed the context in which trade unions operate. They call for tighter regulations, better control and a revision of Lex Laval. Stockholm is also home to "the union centre for undocumented workers" a meeting point where undocumented migrant workers can get information and help they are in a conflict with employers. The centre is managed on a voluntary basis by the three major trade union confederations: Swedish Confederation of Trade Unions (LO), The Swedish Confederation of Professional Employees (TCO) and Swedish Confederation of Professional Associations (SACO). Similar to their Stockholm equivalent, the Building Workers Union in Gothenburg has adjusted to the new context: they now have a Polish speaking ombudsman, and occasionally offer evening courses in trade union history in Polish, and have added courses to their education program specifically aiming to reduce xenophobic attitudes and racism among members. Another example is how the local branch have facilitated enrolment for non-residents, by e.g. accepting cash payments.
On balance then, Sweden proves to be yet another case where multi-level governance is largely lacking when it comes to the primary issues areas in which actors seek to manage the effects of free movement. No single response dominates, as in the other EU-member state cases, and the policies show that various forms of horizontal governance are central to the management efforts. Yet, such a characterization needs to be tempered by an awareness of those responses that suggest both multiple-level and top down forms of governance.

Multilevel Governance
It is abundantly clear that CEE free movement is a policy field with multilevel political system repercussions. The effective implementation of the EU free movement law in the EU Member States and exercising EU free movement rights have not been straightforward. The complexity of a free movement challenge has meant that many policy processes did not involve effective multilevel coordination which would indicate a relationship or participation of all three or four levels of government (European, national, regional and local) in the development of policies, laws or legislation on CEE migration. Instead, traditional models of top-down coordination and localist models have been the most common governance response.
There are however some specific issues that have been handled by applying multi-level governance approach. One such issue is the governance of human trafficking prevention in Sweden and Austria where EU citizens recruited for prostitution come primarily from the CEE countries. There is formal, institutionalised interaction, horizontally and vertically, within an identified, specific policy area. The cooperation ranges from the international organ Interpol, to the operating units in the urban regions, via Europol, police authorities in CEE member states, the National Support Against Prostitution and Human Trafficking, and the regional coordinators. There is a clear regulatory setting and policies are jointly formulated. Another example is the multi-level governance approach to deal with Roma CEE free movers. In Sweden the Network for Roma EU citizens was established. The network connects stakeholders from all over Sweden, public and private as well as nongovernmental, and is well established in both the CEE region and the EU institutions. Based on the EU Framework for National Roma Integration Strategies up to 2020, the Federal Chancellery has adopted an Austrian version in order to ensure Roma inclusion in various domains. From 2012 until 2015, a total of 13 network meetings have taken place including various actors from national and local governments to NGOs and Roma organisations (ATAT-2-3). These meetings aim to encourage an institutionalised dialogue in order to establish Roma specific objectives on the administrative level and to enhance cooperation.
The fact that the free movement is a policy field where the competences are shared in the multilevel setting showed more in the concrete policy processes than in the established governance approaches. The complexity of the free movement implies that the policy authority is located at different levels, and leaving open the possibility of unclear chains of command and a lack of audit leading to the situation where actors expect other actors to take the lead. For example, in the Swedish case the local policy makers required national (or EU) directives in order to handle issues such as registration, begging and homelessness. In the Dutch case, the Minister of Social Affairs and Employment (Asscher) published an opinion article entitled: 'Code Orange for free labour mobility within the EU' to raise European awareness on this topic. He put priority on bilateral collaboration with Poland, Romania, Bulgaria and Croatia (and also on EU lobbying), raising issues like irregular labour contracts and mala fide temporary labour agencies, the adequate provision of information to and equal treatment of border workers. In Sweden, government representatives initiated collaboration between Romanian and Bulgarian politicians, civil servants and civil society representatives in order to discuss the challenges related to Roma beggars in Sweden. The focus was to ensure that these persons would have other solutions than to come to Sweden to beg. Beside horizontal engagement with other member states, the lobbying related to the free movement of CEE citizens was also directed towards the EU itself. For example, in April 2013, four Ministersrepresenting Austria, Germany, the Netherlands and the United Kingdom -wrote a joint letter to the European Commission. In that letter they asked for attention to the need for EU citizens to undergo civic integration, and to consider tightening free movers' rights to social security and to fight fraud and systematic abuse in connection with the freedom of movement more effectively.

Conclusions
In spite of a multilevel setting, not much has emerged in terms of multilevel governance approaches. The EU level is more or less absent. Curry suggests in his contribution to this volume that it can be seen as troubling that little awareness exists among actors involved in free movement management as to why genuine multilevel governance could accomplish, were it to be implemented. Indeed, not just from a policy standpoint, but also from a scholarly perspective, it would be illuminating to hear from those with formal competencies in this issue why the normative preference for multi-level governance represents an actual virtue. Yet, before we can ask -either normatively or theoretically -"why multi-level governance", we need to hammer home that, in this instance, there is essentially no multi-level governance, and then lay the groundwork for understanding why this is this case. Instead, what stands out is that many of the issues connected with CEE migration appear at the local level and are dealt with at the local level. Although local level municipalities and cities may sometimes seek for financial or legal support from the national level a more immediate governance response has been to develop horizontal ties and networks at the local level. As such, the local level responses in this chapter should not be interpreted as representing a bottom-up model of governance. While local horizontal governance networks have taken the lead in structuring governance responses to free movement management, the governance networks have chiefly been that -horizontal. While coordinated responses are indeed also characteristic of the responses in in several of our cases -most notably in the form of multiple-level governance, these cannot be characterized as instances wherein the "bottom" level was decisive. As Curry notes in Chap. 8 of this book, the relationship between local and national levels can best be defined as complex. Hands on solutions involving local actors are rife. However, a distinct feature of the comparison between different cases is that different traditions about constitutional logics set their marks on CEE governance responses as well. Among the selected cases, Turkey is the odd one out in the sense that Turkey is not an EU member and does not apply any regulation about free movement. In Turkey CEE migration governance does not seem to differ much from what is to be expected given normal Turkish top down procedures. Austrian officials resort to regular Austrian corporatism involving tripartite arrangement whereas Sweden and the Netherlands are mixed cases.
CEE migration concerns a number of policy domains but primarily issues that have to do with basic needs dominate: housing, employment, social security. These can be found in all our cases but the governance responses differ to some extent. Inside each domain a number of policy areas dwell and policymakers sometimes differ regarding on which policy areas they have put their attention. While housing is a concern everywhere, Swedish governance actors have been more focused on homelessness than on other housing concerns. The opposite focus can be found among Dutch policymakers. While policies for shelters exist the main concern has been about abusive landlords and about neighbourhood disturbances and to prevent figures of immigrants to exceed certain levels on particular housing estates. Similar differences exist regarding labour market policies. Austrian authorities have put a lot of emphasis on how to avoid nostrification and to help migrants gain necessary educational diplomas. This has not been a major concern in neither the Netherlands nor in Sweden. At this juncture, we will not speculate on why these governance approach differences occur, as within the scope of this project, there is not sufficient data to reach conclusive statements as to causality. This however, does not preclude us from engaging in the time-honoured pastime of academics: informed speculation with the intent of suggesting possible causal relationships that can be more closely examined with supplemental data in the future. It is clear that depending on the policy area, different institutional logics apply in different countries (and sometimes in different regions within the same country), and different historical traits regarding the relationship between government and migration have also developed. Moreover, the character of the CEE migration flows may differ. For instance, the Netherlands have a huge agricultural sector, which attracts large numbers of seasonal workers. Vienna in Austria attracts large number of commuters from neighbouring countries with which there are historical ties as well. This suggests to us that a fruitful strategy for grappling with issues of causality would be one that continues to be multi-disciplinary in nature. Indeed, as historical institutionalists, we argue for -if not the primacy of institutional relationships in shaping governance outcomes -then at least that they will rank among the most important factors. However, the cases examined here suggest that the historical experience of the local and national communities with migration also have much to tell us, and we would consider it beneficial for subsequent analysis to factor in the manner in which collective memories regarding migration are articulated in manners that are often path dependent -at both the organizational level within the public sector, but also outside of the public sector, most notable from the media.
Finally, our emphasis on variation in flows and types of CEE migrants suggests that not only does continued work need to be done on mapping flows of CEE migrants -and especially ensuring comparable data -but that scholars with an interest in free movement management consider how flows and types are made salient in various societal discourses, and how such discourses are then invoked when policy and administrative measures are adopted. Such analyses might also consider the way in which CEE migrants were problematized during the debates over the implementation of transitional arrangements in the early 2000s. As Sweden is the only one of the three not to have made use of transitional arrangements, and as the Swedish responses do not vary substantially from those of Austria or the Netherlands, there is little to suggest that the presence of transitional arrangements exercised causal influence on subsequent governance responses. Yet, this does not exclude the possibility that debates over transitional arrangements left their mark on the way in which governance actors have continued to formulate the challenges associated with free movement management and the related policy, programmatic and administrative tools that have been adopted.
The question remains as to whether an institutional effect is that CEE migration is treated much as any form of migration or as a separate 'EU mobility' case. Obviously, the fact that CEE migrants are EU citizens affects any governance response. Still, a number of Austrian governance responses do not seem to separate between CEE migrants and other migrants, this is particularly evident regarding labour market policies. Swedish officials on the other hand have a wide range of well-established policies that apply to third country nationals while the governance of CEE migrants often seems to fall between two stools.
For decision makers, we conclude with following observation: If those with broad discretion over free movement management -whether at the EU level or among the member states -wish to see the EU involved more via multi-level governance approaches or otherwise, a number of institutional changes will be necessary to consider. One route to go about this would be to create strong policy tools at the EU level concerning housing, labour market and social security. When applied, these tools would create incentives for governance approaches that would involve the EU level; thus multi-level governance responses might develop.
Along an alternative route, some kind of powerful EU legislation would have to emerge. This legislation would have to set clear and unequivocal steering standards, including strong audit institutions, regarding how the member states treat CEE migrants, or indeed EU migrants generally.
Both alternative routes would entail a stronger supranational approach.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.