Multilevel governance from below: how Dutch cities respond to intra-EU mobility

ABSTRACT

‘Intra-EU mobility’ from new member states provides a governance challenge to European countries like the Netherlands. Freedom of movement within the EU enables mobility but also has important social consequences at the urban level in particular. This article discusses to what extent local, national and European governments have interacted in the governance of ‘intra-EU movement’ and how this has affected their policies regarding migrants from Central and Eastern Europe in particular. Focusing on the Dutch case, including the cities of The Hague and Rotterdam, this article shows a development from a decoupled relationship, to localist governance and only recently evidence of emerging ‘multilevel governance’. Speaking to the broader literature on multilevel governance, this article firstly shows that in spite of its broad theoretical application, multilevel governance should be seen as one of the varied types of governance in a multilevel setting. And secondly, it shows how and why local governments can play a key role in the bottom-up development of governance in a multilevel setting.

KEYWORDS:

• Multilevel governance
• local government
• European migration
• free movement
• urban consequences

Introduction

After the EU enlargements of 2004 (Poland, Baltic States and other countries) and 2007 (Rumania, Bulgaria), intra-EU mobility from new member states in Central and Eastern Europe to the ‘old’ EU increased rapidly. In countries like the U.K., Sweden and the Netherlands, intra-EU migrants became one of the most significant immigration categories. The EU principle of free movement thus changed the context of immigration and integration policies at the local and national level in these countries in a profound manner (Balch 2010; Menz and Caviedes 2010; Black et al. 2010; van Puymbroeck, de Pol, and Oosterlynck 2011; Friberg 2013; Caponio and Jones-Correa 2018).

This article discusses the politics of intra-EU mobility in a multilevel setting of the European, national and local level. It questions how local, national and European governments interacted in the governance of Central and Eastern Europe (CEE) migration to the Netherlands, how these interactions can be explained, and to what kind of governance arrangements this has lead. This speaks to the broader literature on multilevel governance, in particular by examining how and why governance structures may or may not emerge in multilevel settings. Whereas most studies on multilevel governance focus on Europeanisation and on the EU-member state nexus (Bache 2008; Hooghe and Marks 2001), this article examines the role of local governments as actors in multilevel settings (Brenner 2004). Intra-EU mobility is chosen as a strategic case study, which includes local, national and European governance levels. To this aim, we adopt a typology of governance arrangements in multilevel settings that focuses in particular on how ‘vertical’ relations between levels of governments can be configured (Scholten 2013). We contribute to existing literature by analysing how and why local governments influence the development of governance arrangements in a multilevel setting.

This article is organised as follows. Firstly, we discuss different perspectives on governance in multilevel settings and outline our research method. Secondly, we analyse how relations between European, national and local governments have been organised and how this has affected the policies at different levels. Thirdly, we analyse these policy practices in terms of the different governance arrangements, finalising with some conclusions.

Governance in multilevel settings

The term ‘governance’ refers to pluri-centric processes that bring together autonomous yet interdependent actors in their efforts to provide solutions to specific issues (van Kersbergen and van Waarden 2004). This can involve horizontally structured arrangements between multiple fields of public, private and societal actors (Rhodes 1997) as well as vertical relations between different government levels. These relationships can be viewed ‘upward’ from nation states towards international institutions, or ‘downward’ from the European to the local level (van Kersbergen and van Waarden 2004; Fenger and Bekkers 2007). As such, governance is conceptualised as steering through a multiplicity of vertical and horizontal relationships.

This vertical dimension of governance, governance in multilevel settings, gained much scholarly attention (Hooghe and Marks 2001; Bache and Flinders 2004; Piattoni 2010). For instance, Hooghe and Marks (2001) paid attention to multilevel governance as an important way of governance in a multilevel setting and distinguished type I and type II modes. Type I refers to how policy competencies are distributed over multiple levels in a clear and formalised division of labour, capturing in particular the formal distribution of competencies between levels of government. Type II refers to more flexible and adaptable modes of interaction between government levels. Especially, Marks and Hooghe’s type II is helpful in the study of how multilevel governance evolves in practice per policy domain or sector.

However, this framework for studying governance in multilevel settings holds two limitations. First, it has been developed (and applied in particular) with reference to European–national relations. For instance, many studies focused on Europeanisation as the key driver behind multilevel governance (Bache 2008), or on the political dimension of nation states towards EU interactions. Consequently, the role of local governments is weakly conceptualised in this framework. This involves not only their role in the implementation of specific policies, but also as actors in processes of agenda setting and policy formulation in such multilevel settings. This is an important limitation against the background of a growing role of local governments in general and cities in particular in global and multilevel systems (Sassen 1999; Holston and Appadurai 1999; Brenner 2004). Therefore, cities should not only be studied as subnational governments that are affected by Europeanisation (Emelianoff 2013) but also as key players in multilevel networks and as motors of policy dynamics ‘from below’.

From this perspective, local governments develop various strategies to influence national and EU policy-making. One strategy involves lobbying or negotiating policy preferences at other levels of government, also described as ‘vertical venue shopping’ (Guiraudon 2000; Pralle 2003). This means that if local governments cannot achieve certain policy preferences at their own level, they can act as policy entrepreneurs in relation with other levels of government. Venue shopping can take place via formal channels where governments of different levels meet, but also via informal policy lobbying or political networks. Such efforts to influence governance processes at other levels also involves networking and coalition building with other cities or local governments and transnational municipal network.

The second limitation of Hooghe and Marks’ type I and type II multilevel governance is that their types always assume meaningful ‘vertical’ relations between various levels. However, the interaction between local, national and European governments as actors in a multilevel setting could cause many different types of governance configurations. Therefore to have an improved understanding of governance in a multilevel setting, this article includes a typology that describes four ideal types of governance in multilevel settings. In this typology, ‘multilevel governance’ is just one probable ideal type that theoretically can be distinguished. Next to multilevel governance, the typology includes centralist, localist and decoupled configurations between governmental levels (Scholten 2013, 2015). These models involve specific forms of interaction between actors at different governmental levels, driven by specific policy and political factors. For instance, some actors may define an issue as a distinct national concern, while others are influenced by local politicisation which affects their governance strategies. Political factors may play a role as well, by specific political party compositions and specific political debates (Breeman, Scholten, and Timmermans 2015). Let us consider these four governance types that we will include in this analysis on CEE migration.

Operationalising types of governance

The centralist type of governance in multilevel settings involves a top-down relationship when it comes to the relation between national and local governments (Sabatier 1986). In terms of vertical relations, this type emphasises hierarchical policy coordination, with virtually no role for local governments in agenda setting and policy formulation. This type is most likely to emerge when there is a clear national political primacy with clear political leadership, when there is a strong national policy coordination structure and when the problem is defined primarily as a national issue. This type holds similarity with the concept of ‘government’ seen from a vertical perspective. However, horizontally, it has attention for the structured arrangements between multiple public, private or societal actors solely active on the national, central level.

This can be contrasted with a localist governance type that refers to governance modes that have (actively of passively) devolved policies to (regional or) local governments. This involves a bottom-up approach where policies are not just implemented but also put on the agenda and formulated by local governments (Sabatier 1986). This perspective is most likely to emerge when problems are defined as ‘local’ and in need of ‘local’ solutions, for instance, phrased as ‘close to the citizen’. Defining issues as distinctly local often requires local leadership on the involved topic, or politicisation at the local level. In this type, local governments are not just policy followers in a hierarchical relationship with central government levels, but they are policy entrepreneurs themselves. It is this type that is most likely to be associated with strategic ‘vertical venue shopping’. Horizontally, it includes structured arrangements between multiple public, private or societal actors primarily active on the local or regional level.

The multilevel governance type distinguishes itself from top-down or localist governance structures by formal or informal vertical interaction between various government levels within a specific policy domain. This includes ‘vertical venues’ for governments from different levels to jointly engage in policy coordination. This can involve institutional fora, regular meetings or networks where organisations from different levels meet. A multilevel governance type is most likely to emerge in situations where the multilevel character of a policy problem is explicitly recognised. This can be the case when problem frames and political frames at different levels are convergent, and can be facilitated by vertical structures or networks. Consequently, specific venues or forums are required for vertical interaction and cooperation. Horizontally, it includes structured arrangements between a variety of public, private or societal actors active on multiple levels.

A fourth type refers to the situation that policies are formulated and implemented on different levels, but without any meaningful interaction. This can be defined as governance decoupling, or the absence of any form of vertical relations or arrangements in multilevel settings. In contrast to the other types, this type marks the absence of joint policy coordination. Evidently, this type can lead to policy conflicts between government levels. Policy paralysis or ‘policy fiascos’ may result when differences between governance levels result in contradictory policies or in the absence of adequate policies (Bovens and ‘t Hart 1996). In Table 1, we operationalised these types of governance arrangements in a multilevel setting primarily by its policy and political elements focusing on horizontal and vertical relationships, interactions and coordination.

Table 1. Operationalisation of governance arrangements in multilevel settings.

	Elements	Centralist governance	Localist governance	Multilevel governance	Governance decoupling
Political elements	Political leadership (by executives) Politicisation of the issue (by political actors)	Political primacy National political leadership Centralist political network	Politicisation at the local level Local leadership Vertical venue shopping	Convergence of political leadership and politicisation Vertical political networks	Disconnected leadership Contested politicisation
Policy elements	Coordination, interactions and relationships of actors and institutions Development of policies, laws and legislation	Central hierarchical policy coordination Issue connections with broader central policies Horizontally structured arrangements with multiple actors active on the central level	Vertical and horizontal policy coordination at local level Little or no central policy coordination Horizontally structured arrangements with multiple actors on the local or regional level	Weak central policy coordination structures Formal and informal vertical structures Horizontally structured arrangements with a variety of actors active on multiple levels	Policy contradictions or conflicts between different levels Absence of coordination and interactions

This operationalisation structures our analysis, including the empirical analysis. It explicitly included policy and political factors to understand governance configurations. This article analyses how and why specific multilevel configurations were developed in the Netherlands in response to intra-EU mobility. We will analyse these configurations in terms of the four types and focus on the role of local governments as actors in the development of these configurations.

Methods

This article involves an in-depth qualitative case study of governance in the Netherlands in response to intra-EU mobility after the EU enlargements of 2004 (Poland, Estonia, Lithuania, Latvia, Czech Republic, Slovakia, Hungary and Slovenia) and 2007 (Bulgaria and Romania). This intra-EU mobility from Central and Eastern Europe is positioned in the institutional context of EU policies and the process of European integration. Therefore, the case of intra-EU mobility was selected as a strategic case because of its distinct multilevel character, relating to the EU principle of free movement as well as to national and local policies of migrant integration. The Dutch case was selected since the Netherlands, like the U.K. and Sweden, experienced a relative large increase of CEE migration since the EU enlargements of 2004 and 2007. The Netherlands has now 120,000 registered and 340,000 estimated CEE migrants (Statistics Netherlands 2012; van der Heijden, Cruyff, and van Gils 2013). In order to capture governance configurations also including local governments, we will focus on two cities in the Netherlands that have received relatively high numbers of CEE migrants: The Hague and Rotterdam. These cities are of comparable size (Rotterdam 607,000 inhabitants and The Hague 510,000 inhabitants) and both have relative large CEE residing and working populations: Rotterdam is now home to between 31,000 and 50,000 estimated CEE migrants, while The Hague accommodates between 17,000 and 33,000 estimated CEE migrants (Municipality Rotterdam 2013; Municipality The Hague 2014). The largest groups are Poles, Bulgarians, Hungarians and Romanians (Statistics Netherlands 2012). In terms of research methods, we used a combination of desk research (key document analysis), semi-structured interviews and a focus group.

The desk research includes key policy memoranda from the European, national and the local governmental levels. All local governmental documents produced by Rotterdam and The Hague, all national documents produced by the Dutch ministries and all European documents produced by the European Commission (EC) in the period 2002–2015 were studied on their importance and relevance on ‘CEE migration’ or ‘EU mobility’. For the analysis of the EU policy context, we analysed key EU policy documents. To select these documents, first the most relevant DGs were selected. By means of the organisational objectives listed at their websites¹ and by a first interpretative study of their policies (a document search on the EC website by specific selection criteria²), a top five of most relevant DGs was prioritised. By this selection process, all official documents of the DG’s Home Affairs, Employment, Justice, Freedom and Security (JLS) and Internal Market were selected online³ by the search words ‘migration’, ‘mobility’ and ‘movement’. By adding all ‘hits’, this resulted in 430 documents which needed an interpretative selection step. By a close reading of these documents’ abstracts, a qualitative indication was provided whether a document was indeed about intra-European movement.⁴ This made it possible to bring down the number of documents to 15 in the end, which were in-depth interpretatively analysed.

On the national level, this was executed by online searches by the number of a dossier.⁵ On the basis of the search on the Dutch translation of the terms ‘mobility’, ‘migration’ and ‘movement’,⁶ four dossier numbers⁷ rose, of which ‘29407’ was the most relevant (‘free movement of employees from new EU member states’). This dossier included 332 documents (2002–2015) which have been interpretatively studied by their abstracts and selected on their qualitative relevancy. This resulted in 53 documents, which contributed to an extensive data file.⁸

On the local level of The Hague and Rotterdam, online searches on the Dutch translation of the terms ‘mobility’, ‘migration’ and ‘movement’ resulted in more than 500 documents (2002–2015) which have been interpretatively studied by their abstracts and selected on their qualitative relevancy. This resulted in 34 documents, mostly consisting of annual ‘migrant monitors’ (‘Monitor EU arbeidsmigratie’), and additional letters, policy communications, laws and legislations which were mostly related to the monitors. In this way, on all levels, a comprehensive and comparative data selection process was completed for the desk analysis.

Furthermore, we included 16 qualitative semi-structured interviews with stakeholders in EU mobility along pre-structured interview guidelines (see Table 2). These were held before and after summer 2015 and included national and municipal policy-makers, lobbyists and private parties. Respondents were primarily local, regional and national civil servants, NGO representatives, private agencies and entrepreneurs active with EU migration and mobility in their daily work. The experts for the interviews were selected by respondent-driven sampling. This holds some clear biases, but as this study focuses on governance approaches, the references and network capital of respondents are important for getting access to authorities and participants in the field. In the respondent selection process, two key criteria were considered: (1) their professional relevance and (2) the geographical dimension. Since we focus on Rotterdam, The Hague and the Dutch case in general, this was an important selection criterion. Next to this, it was important to have respondents who were professionally affiliated to the case in the period of study (also considering public, private and NGO organisations). Table 2 shows the distribution of interview respondents.

Table 2. List of interview respondents.

#	Affiliation	Type
1	Ministry of Internal Affairs	National Public
2	Ministry of Social Affairs and Employment	National Public
3	Ministry of Social Affairs and Employment	National Public
4	Ministry of Social Affairs and Employment	National Public
5	Ministry of Social Affairs and Employment	National Public
6	Ministry of Social Affairs and Employment	National Public
7	Municipality The Hague	Local Public
8	Municipality The Hague	Local Public
9	Polish Embassy	National Public
10	LTO, Agrarian lobby organisation	National employers org
11	OTTO workforce (temporary empl agency)	National Private
12	Municipality Rotterdam	Local Public
13	Municipality Rotterdam	Local Public
14	Polonia	Local NGO
15	Municipality Delft	Local Public
16	Municipality Lansingerland	Local Public

Thirdly, we organised a focus group with 12 stakeholders⁹ in which the preliminary findings of this study were discussed (see Table 3 for the list of participants). The participants were again selected on their professional affiliation (public, private and NGO) and geographical location. They were asked comparable questions as in the semi-structured interviews. These data were also transcribed and analysed by the same codes and subcodes used in the interview analysis. In total, 16 expert interviews and 12 focus group participants are included, adding up to 26 unique stakeholders in this research (Table 2, Table 3).

Table 3. Attendees focus group.

	Organisation	Type
1	Ministry of Internal Affairs	National Public
2	Municipality The Hague	Local Public
3	Ministry of Social Affairs	National Public
4	Municipality Westland	Local Public
5	Municipality Rotterdam	Local Public
6	SNCU	National Private
7	VHA	National Private
8	Barka	National NGO
9	BOOR	Local NGO
10	Obs Kameleon	Local NGO
11	IDHEM	Local NGO
12	IDHEM	Local NGO

Table 4. Periodic overview of Dutch governance arrangements related to ‘CEE migration’.

	2004–2007	2007–2011	>2011
Political elements	Governance decoupling Divergent leadership between local and national level Comparable politicisation on national and EU level concentrating mainly on regulation and implementation of intra-EU mobility	Localist governance Local politicisation Local leadership by aldermen Vertical venue shopping by municipalities Deviating politicisation between local versus national and European level EU mobility paradigm is maintained at EC and national level	Multilevel governance Convergence of political leadership between local aldermen and ministers Convergent politicisation of the issue by Ministers and aldermen
Policy elements	Mainly ad hoc policy measures at the local level Policy contradictions between local and national levels Absence of coordination and interactions between local and national level	Structural policies at the local level Horizontal policy coordination at local level with local stakeholders to solve local concerns Little or no central policy coordination	Institutionalisation of vertical relations between national and local government Coherent national–local policy approach Joint national–local policy lobby towards EU Development of soft-governance framework at EU level

‘Intra-EU mobility’: politics in a multilevel setting

To understand the configuration of relations between levels of government, we first need to contextualise these levels. Especially with the politicisation of CEE migration and the rise of populist discourse in national and local politics, politics matter regarding this issue. In this section, we first explore the context on the EU, national and local level.

EU policies

Migration from Central and Eastern Europe should be positioned in the institutional context of the EU. The European policy context on ‘CEE migration’ focuses on ‘mobility’ derived from the European ‘free movement regime’ (Recchi 2015). This is since the EC perceives ‘free movement’ as the cornerstone of ‘development of the Single Market and the successful integration of the European economy under the Economic and Monetary Union’ (EC 2002, 72, 6). And by Europe as ‘Internal Market’, European citizens can move freely as ‘mobile’ Europeans, not as migrants. This ‘mobility’ perspective contrasts with the movement of ‘Third Country Nationals’ as ‘migration’ of ‘migrants’ from outside the European Union, which needs specified requirements to get access, work and residency in the EU (Favell 2008; Guild and Mantu 2011). Therefore, the ‘movement’ of European ‘mobile’ citizens needs to prosper the development of the ‘Single Market’ stimulated by the EC by removing ‘barriers’, ‘obstacles’ and ‘hindrances’ that frustrate flexibility (EC 2002, 2004, 2007, 4). Therefore, ‘hurdles’, ‘obstacles’, ‘skills shortages and bottlenecks’ need to be combatted (EC 2002, 72, 2013, 837, 2014 , 10), which resulted in policies such as the Europass framework, the MobiliPass and the Job Mobility Portal (EC 2002, 72, 13).

Although national and local governments are free to adopt integration measures on a voluntary basis, mandatory measures or any provision related to entitlements are in this perspective perceived as ‘burdens’ and ‘hindrances’ that illegitimately disturb the EU regime of free movement (Recchi 2015). However, over the past couple of years, the EC has announced several soft-governance measures in relation to the position of EU migrants. In 2013, the EC issued a policy paper on the free movement of people that included a number of actions to help local and national authorities in the member states to regulate new forms of EU migration (EC 2013). With this Action Plan, the Commission acknowledged the difficulties experienced by local governments and took a more active role in supporting local authorities in implementing EU policies.

National policies

Within this European framework, the Netherlands has historically played a foundational role in the construction of the EU, the Euro and has been a proponent of the European free movement regulations. Migration from CEE to the Netherlands did not start with the EU enlargements of 2004 and 2007, but after these enlargements, the number of (officially registered) residents from CEE countries in the Netherlands increased rapidly. In the late 1990s, there were about 50,000 CEE residents, while in 2003, shortly before the EU enlargement of 2004, this number raised to 62,000 CEE residents, now adding up to 200.000 CEE residents (Statistics Netherlands 2011).

In Dutch (national) politics, ‘free movement’ became an important political issue, under pressure of anti-immigrant and anti-EU parties such as the Freedom Party (PVV), in the 2000s. This was in the aftermath of assaults on the politician Pim Fortuyn (2002) and filmmaker Theo van Gogh (2004) that immigration issues became strongly politicised. Therefore, by the accession of the new European member states in 2004, the Netherlands imposed a transitory regime to free movement of workers from CEE countries until 2007. Consequently, Polish citizens needed a work permit until May 2007 and Romanians and Bulgarians citizens until January 2014 to freely work in the Netherlands. These transitional arrangements notwithstanding, in the end numbers more than doubled between 2004 and 2015. This made free movement, the transitional regime and the increasing numbers of CEE citizens in the Netherlands important political and policy issues at the Dutch national level.

The central items on the national policy agenda concerned mainly labour market issues. This includes both illegal practices, like mala fide temporary employment agencies that avoid legal minimum wage, and (semi-) legal practices such as ‘self-employment’ or the ‘posting’ of workers that result in ‘unjust competition’ for Dutch workers (Temporary Parliamentary Commission 2011). Next to this, housing was of primary interest because with a large stock of social housing in the Netherlands, temporary and more permanent housing was a problem. Mobile workers, especially those who have more permanent settlement intentions, face problems with the Dutch housing stock. A final issue is related to registration, since registration procedures are not suited for ‘temporary residents’. Mobile workers are not obliged to register themselves if they stay less than four months in the Netherlands. Therefore, new registration tools were needed to get better insight in ‘temporary migrant’ groups (Ministry of Internal Affairs 2013).

Local policies

From a political historical perspective, Rotterdam has been characterised by a working-class image, continuously led by the Labour Party (PvdA) from the Second World War, which influenced the local political culture. The 2002 local elections marked a turning point in Rotterdam’s politics and administrative culture, described by some as a ‘regime change’ (Tops 2007). These elections were won by the ‘populist’ termed political party (Liveable Rotterdam with 34.7%), which resulted in the first executive board without the Labour Party. Although the Labour Party regained control of the City Council after the 2006 and 2010 elections, Liveable Rotterdam won the elections again in 2014. Next to the dynamic political climate of Rotterdam, this political turmoil needs to be addressed because of the significant importance of issues related to migrants, immigration and integration. From their entrance in 2002, ‘Liveable Rotterdam’ criticised the approach towards migrants, criminality and safety issues. Also in 2014, one of the contested issues concerned the issue of CEE migration since they ‘wanted a quota on the “inflow” of Central and Eastern Europeans’ (Leefbaar Rotterdam 2014, 8). Rotterdam registered 14,883 CEE migrants at the end of 2014 (Municipality Rotterdam 2015). Since many CEE migrants do not register, the municipality of Rotterdam estimated that it hosts between 31,000 and 50,000 CEE residents in 2013 in total and between 18,000 and 38,000 non-registered CEE migrants (Municipality Rotterdam 2014).

The political development of Rotterdam is only partially reflected in The Hague, as it did not have this abrupt political change in the council composition. However, in The Hague, historically led by the Labour Party and the Christen-Democrat party, an anti-immigrant party entered the political stage in 2010; the local branch of the Freedom Party (PVV) entered the City Council. While they had local successes, the Freedom Party did not take office in the municipality of The Hague. In The Hague 16,831 registered CEE migrants resided in 2014, which was a doubling from 2009 (Municipality The Hague 2014).¹⁰ And since many CEE migrants do not register, the municipality estimated that it hosts about 33,000 CEE residents in 2014.

In both cities, there was significant political attention for CEE migration. In The Hague, the Labour Party studied the position of Polish residents, which caught the attention of the City Council (PvdA 2008). Also in Rotterdam, several political parties asked attention for new residents from CEE countries, which resulted in research by the municipality aimed to explore the situation of Polish residents in the city (Municipality Rotterdam 2008a). From that moment on, both cities defined ‘CEE labour migration’ as a local political issue that required monitoring and intervention (Municipality Rotterdam 2008b; Municipality The Hague 2009). Coincidentally or not, the aldermen in both cities responsible for policies on CEE migrants were representatives of the Labour Party (Karakus in Rotterdam and Norder in The Hague).

‘Intra-EU movement’: the policy process in a multilevel setting

After this empirical sketch of problem definitions and the level-specific political and policy context on ‘intra-EU mobility, we will now focus on the governance process and the interaction between these levels. To what extent did the various governmental levels interact, what were their strategies, and how can these vertical interactions be described in terms of the four ideal typical configurations of governance in multilevel settings? We distinguished three periods since the EU enlargement in 2004: decoupling from 2004 to 2007, localist governance from 2007 to 2011 and the gradual emergence of multilevel governance since 2011. We will now reconstruct these phases.

First phase: decoupling

During the period before and shortly after the EU enlargement of 2004, the EC was primarily focused on the enforcement of the free movement principle (EC 2010). The main concern on the EU agenda was the transitional regimes of member states and the removal of restrictions and barriers to ‘finalise’ the Single Market and free movement (EC 2007, 2010). This was comparable with the Dutch national level, since there was elaborate consideration about ‘suspending measures’ in the transitional period (Letter to Parliament 2004). In this period, national discussions focus primarily on labour market permits, labour market tests and the extension or abolishment of restrictions (Letter to Parliament 2004). On the local level, local authorities focused primarily on the most immediate concerns, like housing issues, non-registration and homelessness. Rotterdam and The Hague tried to set up interactions with other large cities such as Amsterdam and Utrecht. This however failed, reflected by a municipal civil servant of The Hague: ‘Utrecht said: “we do not see this problem. We do not have any labour migrants”. And Amsterdam said exactly the same’ (MCS4). Since collaboration with partner municipalities was blocked, Rotterdam and The Hague collaborated and with their own local private agencies and civil society organisations. These local (horizontal) governance relations were mainly focused on proper housing accommodations with housing corporations and labour agencies (such as Vestia, Haaglanden, Jobcenter) to combat illegal housing situations. This was mainly executed by the municipalities of Rotterdam and The Hague and several labour agencies (such as ABU, NBBU and VIA). Consequentially, these municipalities developed all kinds of horizontal relationships with employment agencies, NGOs, the police and social services to tackle issues such as housing controls, labour market controls and safety regulations.

Vertical interactions between the different governmental levels on issues of CEE migrants were largely absent in this period. There were no vertical relations or venues for institutional collaboration or coordination, or as one civil servant from The Hague noted: ‘In that period I felt myself as someone shouting in the desert’ (MCS4). Because there were no coordinated interactions between the national or European level, nor between the local and the national level. The interactions between national and EU institutions were primarily concerned with extensions on the implementation of free movement for CEE workers, while local governments were concerned with ad hoc solutions. This period can be characterised by different problem perceptions, deviating political urgencies and policy concerns. Hence, we refer to the governance mode in this period in terms of decoupling.

Second phase: localist governance

This first phase of decoupling started to change from about 2007. This change was driven by local governments in particular, and triggered an increasing involvement of the national level. Based on horizontal local relations, The Hague and Rotterdam cooperated with local organisations on issues such as housing, labour market and safety. For instance, on social issues, The Hague cooperated with different actors (such as the police, NGOs (IDHEM), schools (Nova College), youth work (The Mall), housing corporations, employment agencies and the Labour Inspection). On homelessness, nuisance and alcohol abuse, The Hague had interactions with several NGOs (Salvation Army, Kessler Foundation, Parnassia, IDHEM) and an employment agency (Otto Workforce). Next to this, Rotterdam signed a covenant on data exchange with the Tax Services and Centre for Work and Income, with the municipality of The Hague, employers, employers interest organisation (ABU, VIA, NBBU) and the unemployment services (UWV) (Municipality Rotterdam 2008b, 2009). Next to this, Rotterdam signed a ‘Covenant housing arrears’ with housing corporations and employers to attack incompatibilities in the housing stock (Municipality Rotterdam 2008b).

One of the first substantial efforts by Rotterdam and The Hague to put ‘CEE migration’ on the national agenda involved a meeting with a wide range of public, private and semi-public stakeholders, the so-called Poles Summit. This ‘summit’ was jointly organised by Rotterdam and The Hague in December 2007. Next to 42 other Dutch municipalities, 2 national Ministers attended the summit. Especially, the responsible Minister of Social Affairs (Donner, Christen Democrats) questioned the concerns of these municipalities since he did not favour extra measures. After this first ‘Poles Summit’, several other summits were held, with the cities of The Hague and Rotterdam drawing further political attention to the ‘integration problems’ concerning ‘CEE migrants’. Notably, the aldermen of both cities took a very critical stance in respect of this inflow of migrants. For example, The Hague’s alderman Norder referred to a ‘tsunami of East European migrants’ (Trouw 2010), while his Rotterdam counterpart Karakus used the Dutch expression ‘mopping the floor with the tap wide open’.

From that moment on, both cities developed a twofold strategy of pragmatic ad hoc solutions on the local level and ‘vertical venue shopping’ in relation to the national level. The acquisition of (‘horizontal’) support from other municipalities was an important element in that new strategy of Rotterdam and The Hague. In this period, Rotterdam and The Hague became close partners, also within the G4 (the association of the four largest Dutch cities). For instance, they cooperated with Amsterdam and Utrecht and implemented policies for homeless people with the NGO of Barka (Polish civil society organisation). This involved non-obligatory efforts to assist homeless ‘mobile EU citizens’ to return to their home country by voluntarily participation. Furthermore, they shared transnational interests with the municipalities of Gent, Antwerp (Belgium), Mannheim and Duisburg (Germany), Margate and Birmingham (the U.K.) within the European networks of Urbiscoop and Eurocities (Municipality Rotterdam 2010).¹¹ In that network, Rotterdam organised the Integrating Cities Conference aimed at more local knowledge sharing and advocated a study into local policies for economic migration within the EU. van Puymbroeck, de Pol, and Oosterlynck (2011, 313) commented: ‘It shows how cities want to engage in knowledge building and exchange “bottom-up”, but at the same time need the cooperation of higher government levels to avoid balancing permanently on the edge of what is legally permissible.’

While these cities broadened their horizontal scope of action with each other and in collaboration with other NGOs and private organisations (such as Barka and temporary labour agencies), they also aimed for vertical interactions. Rotterdam wanted to ‘knock at the door of the national government’ to get things arranged, since they realised ‘we cannot do things on our own’ (Municipality Rotterdam 2008b, 27). Therefore, both aldermen of The Hague and Rotterdam held an intensive lobby at the National Parliament, as one former aldermen noted:

Along the way I went through all chairmen of all political parties in Parliament and described the problem. […] I’ve shown how many people we encountered in those houses, which scared people. Then the government is awakened by the Parliament: you have to do something about this!

The Hague and Rotterdam also proactively invited Dutch Parliamentarians and Euro parliamentarians and guided them through the city and showed them their problems. However, vertical interactions remained limited. While municipalities and Ministers spoke at several occasions about this topic, there seemed to be no shared sense of urgency. Or as one ministerial civil servant indicated this relationship:

We didn’t have much to offer. And our Minister was convinced that these municipalities should not complain so much. […] Next to this he thought that these municipalities were addressing issues that should have been addressed in the past. And that they now used the emergence of new EU citizens to get money from the national government. He didn't want to go along with that. (MCS1)

This stalemate is also confirmed by a municipal civil servant of The Hague, who reported on this period as:

The Ministry of Social Affairs really obstructed any interactions. Really by saying: well, municipality The Hague, do it yourself. Sort it out. We had hard and heavy conversations. […] and I noticed that at that time both Ministries of Social Affairs and Internal Affairs were not cooperating with each other. (MCS2)

It illustrates a period with different political leadership, problem definitions and policy concerns on the national and local level, which did not result in direct vertical interactions. Despite their concerted horizontal efforts, their vertical relationships and policy coordination stayed minimal. Hence, this phase can be seen as a ‘localist’ type of governance in this multilevel setting.

Third phase: towards multilevel governance

The governance interactions changed in 2011 when a new Cabinet was installed and a new Minister of Social Affairs (Kamp, Liberal Party) took office. This new Minister changed the local–national nexus in a significant way, reported by a ministerial civil servant as ‘the tone became different: stricter’ (MCS1) since ‘we got a General on the top. That was a new and confrontational experience’ (MCS2) and importantly ‘Kamp said: “those municipalities were right, we had made a mess of it”’ (MCS3). At the same time, the Dutch national Parliament installed a Temporary Parliamentary Commission ‘Lessons concerning recent labour migration’ (LURA).¹² It was confirmed as a ‘breakthrough’ by multiple civil servants in terms of national awareness on CEE migration issue. This resulted in a welcoming approach towards the municipalities and changes in policy issues related to ‘CEE migration’ (Letter to Parliament 2011). Organisationally, an interdepartmental ‘Taskforce on EU labour migrants’ was installed at the Ministry of Social Affairs which coordinated relationships and interactions between all governmental and non-governmental parties, as reported by a ministerial civil servant:

The ones that shouted the hardest were put on our table. The municipalities of The Hague, Rotterdam and Westland played a very important role in that trajectory […]. Mostly we gave them pre-access in confidential pieces, often they put things on our agenda. […] we really had weekly contact. That was really very intensive. (MCS4)

This leadership change triggered intensive contact between the municipalities of Rotterdam, The Hague, Westland and the Ministries of Social Affairs and Internal Affairs. As a result, as befits the Dutch tradition of corporatism, the Ministry of Social Affairs and Employment developed a threefold collaborative structure: firstly, a managerial group structure (of political representatives); secondly, a steering group structure (of administrative supervisors); and thirdly, a working group structure (of administrative policy-makers) was set up to collaborate and discuss issues regarding CEE migration. These structures had a clear ‘vertical’ character: municipalities (including The Hague and Rotterdam) and various stakeholders were represented in these steering groups. Next to this, steered by The Hague and Rotterdam, the Ministry of Internal Affairs facilitated a ‘Municipalities network’ to discuss local problems and policy initiatives related to CEE migration, joined by 80 municipalities which included 5 annual meetings to share specific knowledge on this topic (Ministry of Internal Affairs 2011). In addition, initially stimulated by private partners, a working group (‘norms-meeting’) was installed on Ministerial level, to discuss housing issues. These ‘vertical’ national–local consultative structures led to a number of national policy measures on the topic of CEE migration, mainly related to registration, work and enforcement of labour regulations, social provisions, (short stay) housing, civic integration and repatriation (Letter to Parliament 2011, 2012).¹³ Through these working groups, municipalities and other stakeholders gained influence in the development of new laws and legislation. As the former aldermen of The Hague adds to this: ‘We delivered complete paragraphs, which were one-to-one taken over […]. We read along, we saw concept Letters to Parliament, we wrote paragraphs. […] we had an enormous influence.’

This resulted in an ‘Intention Declaration’ to enlarge the local governmental instruments (Ministry of Housing, Neighbourhoods and Integration 2012). Next to this, the Ministry of Housing spurred the ‘National Declaration Housing Labour migrants’, which was signed by several municipalities (Rotterdam and The Hague), ministries and corporations that agreed to build extra housing capacity. Furthermore, the Ministry of Social Affairs cooperated with Rotterdam, an employment agency (ABU), a control agency (SNCU) and the Chamber of Commerce Rotterdam to counteract mala fide labour agencies. And finally, to improve registration procedures, the Ministry of Internal Affairs developed a new law ‘Register Non Inhabitants’, applied by 18 (pilot) municipalities to monitor the ‘temporary’ group of CEE migrants who were largely ‘out of sight’.¹⁴

While various interactions were getting institutionalised within the local–national nexus, Rotterdam also developed an explicit strategy regarding ‘Brussels’ (Municipality Rotterdam 2012, 2). The Rotterdam Mayor Aboutaleb issued an opinion paper in the Committee of the Regions to address the local importance and regretted ‘that the Commission’s Social Investment Package neglects the role of local and regional authorities’ (Committee of the Regions 2013, 3). In addition, The Hague and Rotterdam issued a Memorandum of Understanding through the transnational city network of Urbiscoop, with collegial cities in the Benelux area. This instance of vertical venue shopping that was increasingly directed at the EU level was now also joined by Dutch national government. The Minister of Social Affairs and Employment (Asscher) politicised intra-EU mobility and its consequences for receiving countries and cities (Ministry of Social Affairs 2013a). He put priority on bilateral collaboration (with Poland, Romania, Bulgaria and Croatia) and also on EU lobbying, to combat irregular labour constructions and mala fide temporary labour agencies (Letter to Parliament 2013, 6).

In this regard, in 2012 the Minister co-organised a conference together with the German Ministry of Interior Affairs¹⁵ and he co-authored an opinion article in The Independent and the Dutch daily De Volkskrant (Asscher and Goodhart 2013). In this article, the Minister drew attention to the downsides of CEE labour migration and addressed the issue of European awareness on this topic. This article was followed by a ‘EU labour migration Summit’ in September 2013 in The Hague, organised by the Ministry of Social Affairs and Employment together with the local governments of Rotterdam and The Hague. This summit was attended, next to many Dutch municipalities, employment and housing agencies, also by the Romanian Minister of Social Affairs and the Bulgarian vice-minister of Labour. At this summit, especially the bilateral and national efforts of cooperation and collaboration (to combat fraud, mala fide employers and slum landlords) were discussed.

In addition, more horizontal coordinated interactions became visible on the national level. First, in 2013, the Dutch minister of Immigration (Teeven), together with colleague ministers of Germany, Austria and the U.K., asked the EC to consider the need for EU citizens to undergo civic integration. They asked for more discretion in terms of tightening migrants’ rights to social security and to fight fraud and systematic abuse effectively (Mikl-Leitner et al. 2013). Besides, the Dutch Minister of Social Affairs (Lodewijck Asscher) signed a letter together with colleague ministers of Sweden, Belgium, France, Austria, Germany and Luxemburg to make adjustments to the Posting of Workers Directive (Ministry of Social Affairs 2012). All these efforts illustrate the importance of horizontal relationships between national governments to aim for vertical interactions with the EU.

In response, the EC proposed an explication of existing legislation and promoted the exchange of best practices at the local level in particular (EC 2013; Letter to Upper House 2013). Therefore, vertical interactions on EU mobility or CEE migration between member states and EC have a different institutionalised outlook than the local–national interactions described above. Communication between the Dutch national government and the EC is more irregular, indicated by a ministerial civil servant:

No, working groups do not exist. […] such meetings that we had with the Commission, I would not say these are institutionalized meetings. It is an ad-hoc meeting. And it establishes several times, as much as needed. But the Commission is not the authority who structurally wants to broach such meetings with several member states. (MCS3)

In sum, in this period we see nascent multilevel governance, mainly regarding multiple horizontal intergovernmental interactions linked to intensive vertical interactions between the national and local level, and with the European level to a limited extent. After existing local horizontal interactions, also the Dutch national government coordinated horizontal interactions between ministries, other EU member states and stakeholders. Efforts were made for ‘horizontal’ coalition building with the aim of raising ‘vertical’ issue awareness at the EU. While the national–European vertical interactions seem to be limited, this period clearly shows the establishment and institutionalisation of local–national interactions, hence a multilevel governance arrangement in this third phase.

Conclusion

This article displays that a nascent multilevel governance structure on ‘intra-EU movement’ was not so much established top-down by European institutions or member states, but rather ‘bottom-up’ by coordinated efforts by local governments. Speaking to the literature on multilevel governance, this shows the relevance of the agency of local governments in general and cities in particular in the development of multilevel governance. Our analysis displays the importance of local governments in the establishment of horizontal and vertical governance configurations in multilevel settings. Local governments frequently engage in ‘vertical venue shopping’ in an effort to influence agenda setting and policy formulation at other levels of government. As our analysis unfolds, this can lead to the establishment of more permanent ‘vertical’ structures that facilitate collaboration across governmental levels. Also, we found examples in which local governments first built ‘horizontal’ coalitions with other local public, private or NGO partners, in order to broaden their scope of action towards ‘vertical’ relations with the national or EU level.

Our case study of how Dutch cities (Rotterdam and The Hague) responded to the increased ‘intra-EU movement’ clearly shows the importance of local entrepreneurship in the establishment of vertical governance relations in a multilevel setting. This entrepreneurship was driven by various factors: a specific problem perception (as CEE migrants settle in specific cities first, local consequences became manifestly visible), policy concerns (on labour market issues, housing and registration) and political leadership (by aldermen driven by clear political agendas). This entrepreneurship needs to be understood in its political context, since this issue caught attention when intra-EU mobility became politicised by ‘populist’ and more mainstream political parties in both cities and National Parliament. In Rotterdam and The Hague, there was broad political consensus to intervene on this issue, put on the agenda by political parties and studies promoted by the City Council (PvdA 2008; Municipality Rotterdam 2008a). In the same way, after broad political consensus, the Temporary Parliamentary Commission put the issue on the national agenda, which under a new Cabinet, made it legitimate for the new Ministers to react on this issue. Our analysis thus shows the importance of politicisation and the significance of political actors in the changing establishment of governance arrangements between levels throughout time.

Strategies to influence policy-making at other levels of government included various forms of ‘vertical venue shopping’. In our case, we found four key steps of how local governments strategically organised agenda setting at the national and later at the EU level. Firstly, agenda setting was launched by active media attention (the aldermen of The Hague calling for a ‘tsunami’ and the Minister for ‘Code Orange’). Secondly, this was continued by broadening the scope of action with horizontal public and private partners (Rotterdam collaborated with The Hague and Westland, temporary labour agencies and NGOs while the Dutch government collaborated with Dutch civil organisations, Sweden, the U.K. and Germany). Thirdly, this was followed up by the organisation of a political event to influence the topic (‘Poles Summits’ and the ‘EU Summit’). And fourthly, in the reorganisation of administrative efforts, the topic got institutional urgency. In Rotterdam, The Hague and the Dutch Ministry of Social Affairs, integral ‘Taskforces on EU labour migration’ were installed. Such efforts by the two Dutch local governments have not been without effect. Incrementally, vertical multilevel governance structures emerged, mainly with the national level and other stakeholders.

This article speaks to the broader literature on multilevel governance, raising attention to the agency of local (as well as national and European) governments in establishing horizontal and vertical governance configurations in a multilevel setting (Hooghe and Marks 2001; Bache and Flinders 2004; Piattoni 2010). Rather than taking multilevel governance as a given, or as ‘independent variable’, our study shows why one cannot assume ‘multilevel governance’ but it has relevance to study this as one of more ideal typical configurations. It displays that a better understanding is required how and why multilevel governance evolves under specific circumstances, as a ‘dependent variable’. This contributes to a more precise theoretical understanding and conceptualisation of multilevel governance. In fact, in our analysis we only found some evidence of institutionalised vertical relations that can be termed as ‘multilevel governance’ in the most recent stage of policies regarding ‘intra-EU movement’. In earlier stages, despite a multiplicity of horizontal relationships on the local governance level, vertical relations were either absent or only ad hoc. In this respect, an important lesson from our analysis is that in spite of its broad theoretical definition and application in the literature, multilevel governance is hard to achieve in practice and need to be seen as one of the varied types of governance in a multilevel setting.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 http://ec.europa.eu/about/ds_en.htm.

2 Document selection on the basis of ‘department’ (the five most relevant DG's selected in the previous stage);‘time’: 1 January 2002 until 1 January 2014; ‘document type’: all; and ‘language’: ‘English’.

3 By ec.europa.eu website.

4 To illustrate this, for instance, a lot of documents regarding ‘movement’ in the area of infrastructure, climate change or maritime affairs could get filtered out by this procedure.

5 Letters to Parliament (Brieven aan de Tweede Kamer) by www.tweedekamer.nl and https://zoek.officielebekendmakingen.nl.

6 ‘mobiliteit’, ‘migratie’ and ‘verkeer’.

7 32680; 29407; 83432; 29911.

8 This resulted in a more than 400-page dossier.

9 This took place on Friday 13 November 2015.

10 As context: the annual growth of the CEE population in The Hague was plus 20% between 2009 and 2012, and 7% in 2014.

11 Next to this, in 2012, the Rotterdam aldermen Karakus, the mayor, the head police officer and the public prosecutor had a work visit to Romania and Bulgaria and visited several places to see the local context of where migrants come from.

12 The effectiveness of the local imagery on this national commission can be illustrated by the beginning of the rapport: ‘[…] images dominated like a tsunami of Eastern Europeans, uncontrollability of inflows and modern slavery in the Netherlands’ (Temporary Parliamentary Commission 2011, 5).

13 Such as for instance: a Toolkit for effective estimations, a new registration system RNI, and adjustments to the ‘Act Exceptional Measures for Urban Problems’ to combat slum landlords.

14 Next to this, a revision on the ‘Act Exceptional Measures for Urban Problems’ (also called ‘the Rotterdam Act’) was introduced on request of Rotterdam. This act enables a selective settlement policy in vulnerable urban districts: low income households are not allowed to settle there. After an intensive lobby by Rotterdam, the Act was revised in 2013, the so-called Rotterdam Act II. The revision was a direct response to Rotterdam's concerns about housing problems of ‘CEE migrants’, providing the local government instruments to combat irregular landlords as well as to ban persons that have disturbed public order from living in specific neighbourhoods.

15 Called ‘free movement and participation of EU citizens – making it work for all’.

Appendices

Appendix A1. All letters to parliament.

Author	Title	Document number/date
Vice-minister Social Affairs Rutte	Free movement labourers from new EU member states/extension of the EU: advantages and disadvantages of a direct free movement of labourers	29407 (1) 23-01-2004
Vice-minister Social Affairs Rutte	Free movement labourers from new EU member states	29407 (8) 13-02-2004
Minister Social Affairs De Geus	Free movement labourers from new EU member states	29407 (14) 29-04-2004
–	–	29407 (16) 02-07-2004
Vice-minister Social Affairs Van Hoof	–	29407 (17) 08-07-2004
Vice-minister Social Affairs Van Hoof	–	29407 (17 reprint) 08-07-2004
–	–	29407 (21) 21-05-2005
–	–	29407 (31) 15-02-2006
–	–	29407 (32) 31-03-2006
–	–	29407 (33) 20-04-2006
–	–	29407 (44) 25-04-2006
–	–	29407 (51) 15-09-2006
–	–	29407 (53) 10-10-2006
–	–	29407 (54) 28-11-2006
–	–	29407 (56) 12-12-2006
Minister Social Affairs Donner	–	29407 (70) 24-04-2007
–	–	29407/ 30678 (72) 16-05-2007
–	–	29407 (73) 14-08-2007
–	–	29407 (75) 27-11-2007
Minister Social Affairs Donner and Housing, neighbourhoods and integration Vogelaar	–	29407 (81) 16-06-2008
Minister Housing, neighbourhoods and integration Vogelaar	–	29407 (95) 30-09-2008
Minister Social Affairs Donner	–	29407 (96) 09-10-2008
–	–	29407 (97) 03-11-2008
Minister Social Affairs Donner and Housing, neighbourhoods and integration Van der Laan	–	29407 (98) 28-11-2008
Minister Social Affairs Donner	–	29407 (99) 11-12-2008
Minister of Housing, neighbourhoods and integration Van der Laan	–	29407 (103) 23-11-2009
Minister Social Affairs Donner	–	29407 (105) 23-03-2010
Minister of Housing, neighbourhoods and integration Van der Laan	–	29407 (106) 25-06-2010
Minister Social Affairs Donner	–	29407 (108) 28-09-2010
Minister Social Affairs Kamp	–	29407 (116) 01-03-2011
–	–	29407 (118) 14-04-2011
Minister Internal Affairs Donner	–	29407 (130) 15-09-2011
Minister Social Affairs Kamp	–	29407 (132) 18-11-2011
–	–	29407 (141) 23-03-2012
Minister Internal Affairs Spies	–	29407 (146) 11-04-2012
Minister Social Affairs Kamp	–	29407 (149) 28-08-2012
Minister Internal Affairs Spies	–	29407 (150) 25-09-2012
Minister Immigration, Integration and asylum/ Leers	–	29407 (151) 02-11-2012
Minister Social Affairs Asscher	–	29407 (153) 14-01-2013
Minister of Housing Blok	–	29407 (154) 18-01-2013
Minister Social Affairs Asscher	–	29407 (162) 07-03-2013
Vice-minister Safety and Justice Teeven	–	29407 (171) 05-07-2013
Minister Social Affairs Asscher	–	29407 (172) 10-03-2013
–	–	29407 (174) 09-09-2013
–	–	29407 (177) 16-10-2013
Minister Social Affairs Asscher and Vice-minister Safety and Justice Teeven	–	29407 (180) 03-12-2013
Minister Social Affairs Asscher	–	29407 (181) 02-12-2013
–	–	29407 (187) 11-02-2014
Minister of Housing Blok	–	29407 (188) 27-02-2014
Parliamentary Commission Lessons recent labour migration	List of questions and answers	32680 (10) 20-10-2011

Appendix A2. Interview template.

Network

1. Spontaneously mention the stakeholder you cooperate with in governance of CEE migration.

Table

Interaction
2. Dimensions of interactions
Question	Operationalisation
Nature of interaction	Formalised/institutionalised	Informal/ad hoc
Frequency of interaction	Regular	Irregular
Content and purpose of interaction	Information exchange/resource exchange/cooperation/alliance/mutual issue interest
Bureaucratic structure
3. Aspects of bureaucratic structure
Question	Operationalisation
Discretion in the work related to CEE citizens	Autonomous	Hierarchical
The decision procedure(s)	Regular	Particular
Financial discretion	Unregulated	Regulated
Multi stakeholder evaluation
4. Evaluation of measures
Do you see the existing measures as effective or not?	Yes, because … .	No, because … .
Where are more effects needed?
Political aspects
5. Political context
Presence of EU sceptic political parties or networks (coalitions)	EU, national, regional, local level
Presence of immigration sceptic parties
Nature of the media attention directed to issues related to CEE migration	High Generally positive	Low Generally negative