Balancing informality and need – policy responses to informal East-West migration in Austria, the Netherlands and Sweden

ABSTRACT

The article provides an analysis of policy responses to mobile EU citizens without legal residence in the Netherlands, Austria and Sweden. A comparative case study design uses qualitative survey and interview data to identify national and local policy responses to the implications of EU citizens from Central and Eastern Europe (CEE) living without legal residence in the aforementioned countries. The theoretical framework specifies how the institutional logic of welfare regimes is likely to generate policy responses that address need, sanction informality or do both. The choice indicates the priority given to redistributive outcomes, administrative procedures, or both. The results reveal similar responses to those implications relating to the labour market, but slightly different approaches to the implications in the social domain. Policy responses to labour market implications have predominantly focused on sanctioning informality. Swedish and Austrian policy responses in the social domain have focused on addressing need. Dutch policy responses to social issues instead focused on sanctioning informality, prioritizing procedure. The results indicate that local and national policy responses to implications of informal intra-EU migration may be fruitfully understood through the prism of welfare regimes and related approaches to need and informality.

KEYWORDS:

• Intra-EU mobility
• social policy
• labour market policy
• welfare state
• informality

Introduction

Since the EU enlargement in 2004, questions about how increasing East-West mobility may threaten the welfare systems and labour standards of receiving member states have been high on the political agenda (e.g. Crespy and Menz 2015; Kahanec and Pytliková 2017; Menz 2003; Mikl- Leitner et al. 2013). Political and judicial responses to these concerns have intensified the stratification of rights among mobile EU citizens, most notably between EU citizens that fulfil the conditions for legal residence and those that do not (Giubboni 2015; O’Brien 2016). The latter category includes numerous EU citizens active in the informal economy. Although not entitled to social assistance or access to labour market policies, their presence has social and economic implications – e.g. exploitation, homelessness, addiction and deteriorating health – in both local and national policy contexts (Garapich 2014; Likic-Brboric, Slavnic, and Woolfson 2013; van Ostaijen, Reeger, and Zelano 2017). As shown by Menz (2003, 2011) and others, policy outcomes result from varying types of interaction between national institutions and the EU level. Although not related to the degree of European integration as such, governance EU citizens without legal residence is constrained by national and local interpretations of regulations at the EU level. According to EU case-law, the lack of legal residence excludes a person from national welfare systems, but do not automatically motivate expulsion. Consequently, member states are left to manage the implications of this group on their own, motivating the analytical focus on national institutions in this article.

This article compares local and national policy responses to the presence of EU citizens from Central and Eastern Europe (CEE) without legal residence in three popular destination countries: Austria, the Netherlands and Sweden. A set of similar, but not identical, welfare state institutions and corporatist labour markets paves the way for interesting comparisons. The article aims to theorize and understand the interaction between non-selective, encompassing welfare regimes and policy outcomes in the governance of EU citizens without legal residence. The research question reads: What policy responses have emerged to manage the implications of CEE citizens without legal residence in Austria, the Netherlands and Sweden? The article aspires to enhance the knowledge of policy making in encompassing welfare states. Previous literature establishes these welfare regimes tend to be more generous towards immigration than liberal ones. Comparing policy responses to EU citizens without legal residence may further disentangle the mechanism behind the low levels of welfare chauvinism. Methodologically, it conceptualizes the policy aspect of governance and applies this empirically. The empirical contribution consists in highlighting the policy implications of informal EU mobility.

The first section defines the group of EU citizens without legal residence and characterizes their situation as one of informality and need. Subsequent sections introduce the theoretical framework, method, data and sketch expectations of the empirical results. Results are reported thematically and are followed by a comparative analysis. A concluding section summarizes the findings and identifies avenues for future research.

EU citizens without legal residence

Several scholars acknowledge a divide between informal and formal labour in the context of free movement, particularly since the Eastern enlargements. CEE citizens are overrepresented among informal workers in the West (Hazans 2011), causing Favell (2008) to caution about the emergence of an informal underclass. Recent political (Crespy and Menz 2015) and legal (O’Brien 2016) developments at the EU level emphasize this distinction. The cases of Brey, ¹ Dano ² and Alimanovic ³ are illustrative examples from European case-law. The judgements break with the previous position of the European Court of Justice (ECJ), considering EU citizenship the basis for access to social assistance under the principle of equal treatment (Hailbronner 2005).⁴ In Brey, the ECJ instead determines that Directive 2004/38 set the limits of Treaty rights, increasing the focus on self-sufficiency. The Dano case goes further still by explicitly stating that the principle of equal treatment only applies if the conditions for legal residence laid down in Directive 2004/38 are met, i.e. sufficient resources and comprehensive sickness insurance (Giubboni 2015). Member States may exclude economically inactive EU citizens from non-contributory social benefits on the grounds of not meeting the conditions for legal residence (Giubboni 2015; O’Brien 2016). In addition, local and national definitions of “economic activity” and “sufficient resources” are becoming more restrictive (O’Brien 2016). The Alimanovic ruling establishes that benefits aiming “to cover the minimum subsistence costs necessary to lead a life in keeping with human dignity” cannot be considered as facilitating job market access. By definition, they then become unattainable for economically inactive EU citizens (O’Brien 2016, 947).

Numerous EU citizens reside in another member state without fulfilling the conditions for legal residence. A heterogeneous group of trained craftsmen, low-skilled or illiterate persons, they work in informal or semi-formal employment or make money from begging or playing music in the streets (e.g. Adriaenssens and Hendrickx 2011; Garapich 2008; Likic-Brboric, Slavnic, and Woolfson 2013; Sert 2014). Their situation is characterized by informality in residence and economic activity. The associated social risk essentially implies individual need. They lack the social and welfare rights of legally residing EU citizens or native residents and make their living from precarious activities. But failing to meet the conditions does not automatically motivate expulsion (Giubboni 2015; Hailbronner 2005). This is only possible if the EU citizen considered to constitute a threat to public safety or health, or that are considered an “unreasonable burden” for national welfare systems (Erhag 2016; Giubboni 2015). The latter criterium, in particular, has proven difficult to apply in practice, as one person is unlikely to affect the economy of an entire member state (Erhag 2016).

In the absence of EU regulations about how to manage their presence, the implications of informal EU mobility become matters of national and local competence. To analyse host society reactions, it is therefore required to build a theoretical framework based on research studying how welfare institutions and labour markets respond to immigration, informality and need.

Institutional logic, informality and need

The framework theorizes a link between national institutions and policy responses to informal EU mobility. It specifically focuses on institutionally conditioned approaches to elements characterizing the situation of EU citizens without legal residence: i.e. need and informality. Previous research finds that the design of welfare institutions generates particular logics; encouraging certain behaviours, values and policies while discouraging others (e.g. Jordan 2013; Pfau- Effinger 2005; Rothstein 1998). Exactly why responses to e.g. immigration vary, is less straightforward. This article adds a piece to the puzzle, by comparing policy responses to informal intra-EU mobility in encompassing welfare states.

The Esping-Andersen (1990, 2009) seminal categorization of welfare states clusters countries into three ideal regime types. Although the typology has been both criticized and extended (e.g. Ferrera 1996; Lewis 1992), the initial distinction between Liberal/ Selective regimes, Conservative/ Christian Democratic regimes and Universal/ Social Democratic ones is still widely recognized. The Liberal regime emphasizes selectivity to a higher degree than the Conservative and Universal regime. In the EU, the UK and Ireland are the only countries habitually categorized as liberal welfare regimes (van Der Waal, de Koster, and van Oorschot 2013). Most EU countries approximate the other types, with similar mechanisms motivating redistribution (e.g. Jensen and Svendsen 2011). While the family and civil society actors (e.g. churches) are important sources of service provision in conservative regimes, the state is the primary service provider in universal ones (Esping-Andersen 1990). The theoretical framework argues that encompassing welfare states have an instrumental as well as a moral rationale to address need and sanction informality.

All welfare states provide citizens with an insurance against social risk and ultimately, need. Such insurance depends on the economic redistribution of public revenue (i.e. taxes, social contributions) derived from formal economic activity in markets (Edlund and Johansson Sevä 2013; Rothstein 1998). Informal work constitutes an instrumental threat to the sustainability of the redistributive system. There are also non-instrumental reasons to safeguard a formal economy and address need. According to Rothstein (1994, 1998), welfare institutions need to be perceived as just in order to generate and maintain public support. Such trust is achieved by redistributing in a just way (procedural justice) with just results (redistributive justice) and by coming across as preventing free riding. To signal procedural justice, public institutions (e.g. tax agencies) need to be perceived as impartial. To signal redistributive justice, institutions must be perceived as delivering fair outcomes (e.g. reduce inequality in accordance with prevailing norms). Public institutions must also come across as safeguarding just procedures by, for example, sanctioning tax evasion.

The functional and moral mechanisms are mutually reinforcing: without public support, no tax revenues and no redistribution, without redistribution, no perceptions of just redistributive outcomes, no tax revenues and so on. The combination described incentivizes the state to sanction informal economic activity and redistribute fairly.

In order to achieve the latter, it is not enough to address need as such. It must be done in accordance with general perceptions of “deservingness” (Daniele and Geys 2015; Larsen 2008; Slothuus 2007). According to Larsen (2008, 149, based on van Oorschot 2000) people may assess deservingness using up to five criteria: (1) Control: the less in control of neediness, the higher the degree of deservingness, (2) Need: the greater the level of need, the higher the deservingness, (3) Identity: the higher the degree of group belonging, the higher the deservingness, (4) Attitude: the more grateful, docile and compliant, the higher the deservingness and (5) Reciprocity: the higher the level of previous or future payback, the higher the deservingness.

Immigrants tend to be perceived as less deserving than other exposed groups, e.g. the unemployed (van Oorschot 2006). The view that immigrants are less entitled to welfare benefits and services than natives and should only partially, or not at all, be integrated into welfare state programmes is captured by the concept of welfare chauvinism (Jørgensen and Thomsen 2016; Mewes and Mau 2013). Universal welfare regimes show lower levels of welfare chauvinism than conservative regimes, and selective ones in particular (Reeskens and van Oorschot 2013; van Der Waal, de Koster, and van Oorschot 2013). The literature offers several explanations. An established explanation is that universal institutions foster tolerance and solidarity to a greater extent than selective institutions (e.g. Boräng 2015; Crepaz and Damron 2009; Kirchner, Freitag, and Rapp 2011), reflected in policies and attitudes attentive to individual need. Another potential explanation is that welfare regimes tend to favour different types of immigration, more or less likely to be perceived as an economic threat. In universal regimes, strong labour market insider interests, i.e. trade unions, tend to actively oppose labour immigration policies as they increase labour supply and competition (Bucken-Knapp 2009; Rueda 2005). Instead, universal welfare states have historically favoured humanitarian migrants and refugees. As refugees often face barriers to employment (e.g. Friedberg 2000; Lundborg 2013), they are less likely than labour migrants to be perceived as a threat to native employment, hence dampening anti-immigration attitudes (Kunovich 2013). According to this view, the lower perceived threat to labour market insiders explains the comparatively low levels of chauvinism in universal regimes, rather than institutionally fostered solidarity towards strangers. So far, the predominance of one type of immigration has made it difficult to disentangle the underlying mechanisms further: Humanitarian migrants are by definition individuals in need with limited opportunities to threaten insider labour market interests. However, EU citizens without legal residence simultaneously represent increased labour supply, potentially pushing down wages and standards, and individual need, existing outside legal and social safety nets. Studying policy responses to a group potentially alluding to both mechanisms, could offer new insights to the institutional logics of non- selective regimes. In order to analyse policy outcomes relating to EU citizens without legal residence, the next section conceptualizes the policy process.

Conceptualizing the policy process

Political and administrative reactions to implications of EU citizens without legal residence are examples of governance. Governance defines the structures within which stakeholders from various sectors and administrative levels interact to direct public resources towards a collective goal (Carmel 2011; European Union 2001; Hooghe and Marks 2001; Piattoni 2010). Governance also connotes policy processes evolving within those structures as stakeholders mobilize to address specific issues (e.g. Dekker et al. 2015; Goodwin 2015; Scholten 2016; Strünck 2005). This article focuses on the second aspect and conceptualizes the policy process as composed of one discursive and one applied, practical element (e.g. Cheung 2011).

In the first phase, stakeholders define what issues to put on the agenda. According to Bacchi (2009), this is where “problematizations” or problem definitions emerge (Cook and Natalier 2013). Problematizations vary with circumstance, e.g. ideological trends or institutional settings (e.g. Béland 2009). Problem definition emerges in various practices, providing a conceptual link between the first phase and the second part of the policy process. Legislation, guidelines, working groups or networks are examples of practices (Börzel and Panke 2007; Dahlström, Peters, and Pierre 2011; Rhodes 2007). Figure 1 illustrates the policy process.

Figure 1. Conceptualization of the policy process.

Method and research design

The case selection includes local and national levels of governance in Austria, the Netherlands and Sweden. They are popular destinations for mobile CEE citizens, some of whom work and reside without legal residence. The similar, but not identical, welfare and labour market structures make it both possible and interesting to compare these countries. Most welfare states in the EU share more traits with universal and conservative regimes than selective regimes of e.g. the US and the UK. A comparison also addresses the puzzle of why encompassing regimes show comparatively low levels of welfare chauvinism. Analysing policy responses to EU citizens without legal residence offers insights to whether non-selective institutions foster policy responses giving primacy to solidarity, formality, or both.

Austria, the Netherlands and Sweden are encompassing welfare regimes but the specific configuration of welfare institutions varies some. Austria is a conservative welfare state and Sweden a universal one. The Netherlands is generally considered a “hybrid” regime, mixing universal and conservative traits (Arts and Gelissen 2002; van Hooren and Becker 2012). The labour markets in the case selection are corporatist with cross-national variations (Esping-Andersen 1990; Hartog, Leuven, and Teulings 2002). The Austrian labour market is characterized by institutionalized tripartite dialogue between labour, business and government(s) (Afonso 2013). The Netherlands combines sector or industry-specific bargaining regimes with a tripartite political dialogue, trying to balance the power of insiders and facilitate reform (Arts and Gelissen 2002; Yerkes 2014). In Sweden, the state is absent from social partner negotiations but supplies complementary policies (Marshall 1996).

An evident difference between the countries is the variation in politico-administrative design, ranging from federalist Austria to the more unitary Netherlands and Sweden (Lijphart 2012). While these differences are likely to influence the structure of governance, i.e. which level or actor does what, rather than the issue of central interest: policy problem definition and resulting practices. They are therefore not considered to hamper the cross-case comparison. Figure 2 provides an overview of the national contexts.

Figure 2. Case selection overview.

Data

Data originate from a stakeholder-centred methodology in two sequences corresponding to the phases of the policy process. Local level data were collected from two urban regions in each country.⁵ To identify problem definitions, an online survey (n = 129) was coupled with 42 stakeholder interviews at local and national governance levels. The respondents worked with issues relating to CEE migration, including informal migration. The data used to identify policies and practices consist of primary and secondary sources relating to intra-EU mobility in local and national jurisdictions, and 52 semi-structured interviews.⁶ Each part of the data collection process closed with a focus group (Urban Living Lab) (Reeger and Enengel 2015; Zelano et al. 2016). Table 1 summarizes the data collection per national context.

Table 1. Data collection.

Phase		Data collection method	Austria	The Netherlands	Sweden
Policy process	Phase 1. Problem definitions	Online Survey	n = 46	n = 46	n = 37
		Interviews	n = 14	n = 12	n = 17
		Focus groups	n = 8	n = 16	n = 30
	Phase 2. Policies and practices	Policy mapping
		Interviews	n = 15	n = 16	n = 21
		Focus groups	n = 15	n = 13	n = 8

An initial analysis motivated the subsequent methodological choices. Problem definitions identified in one national context only were excluded.⁷ The risk of losing cross-context variation was considered lower than the risk of losing analytical depth as a result of including a higher number of problem definitions. Problem definitions relating to the market for sexual services were excluded as the radically different approaches to the sex trade made a meaningful comparison unsustainable.

Operationalizations

The first phase of the policy process (problem definition) is operationalized as the implications of informal CEE migration as highlighted by stakeholders. The second phase (policies and practices) is operationalized as written policies and/or repeated actions on the national and local level addressing such implications. Informality is operationalized as living and/or working outside formal registers, without formal employment or housing contracts or a registered right to residence. Need is operationalized as a legal, material, medical or psychological condition inferior to the institutionally defined minimum in a specific national context.

Some expectations

Member states manage the implications of mobile EU citizens without legal residence within an institutional setting that encourages behaviour that addresses need and sanctions informality. As the situation of EU citizens without legal residence is characterized by both need and informality, the situation holds a potential dilemma. The empirical analysis aims to describe how stakeholders in host societies manage this. Stakeholders may focus on addressing need, sanctioning informality or both. Policy responses catering to individual need, is taken to indicate a solidarity based reaction favouring fair redistributive outcomes. Policy responses prioritizing the formality dimension, rather indicate a preference for procedural justice. The priority could vary across the different phases of the policy process, resulting in four possible combinations of problem definitions and practices. They are illustrated in Figure 3.

Figure 3. Possible policy responses to informal CEE migration.

A general expectation is that the stronger the concentrated interests in the labour market, the more likely it is that sanctioning informality (and limiting competition) is prioritized over addressing need (e.g. Pierson 2001). Assuming a presence of insider interests in corporatist labour markets, Outcomes B, C and D are likely responses to labour market implications (cf. Rueda 2005). Whether or not the different models of corporatism result in different outcomes remains to be seen.

Policy responses to issues outside of the labour market are harder to foresee. The deservingness criteria (above) nonetheless serve as guidance for reflection. Assessments relative to the criterion of Identity may go either way: informal CEE citizens may be considered as belonging (i.e. fellow EU citizens) or as outsiders (i.e. non- citizens). The criterion of Reciprocity (future, present or past) is not likely to be perceived as fulfilled, primarily because the group do not pay tax or contribute to the formal economy. Assessments regarding the third criterion, Control, are likely to depend on assumptions about the underlying causes for being in the host member state. Discrimination in countries of origin or income differences, might be considered to reduce EU migrant’s level of control over their situation. The opposite applies if their presence is perceived as a manifestation of fraud and non-compliance. The criterion Attitude also allows for various outcomes. Informal EU citizens may be perceived as “docile” and “compliant” as they do not claim benefits and exist under the radar of host authorities. But they may also be perceived as non-compliant as they lack legal residence. Considering the ambiguity, it is not possible to predict which way policy responses will swing. It is, however, possible to say that Outcome A is more likely to occur in cases of very tangible need, such as visible begging or homelessness, than in cases where the informality aspect is not as obvious.

Case-specific expectations are limited. As labour market insiders are differently organized, varying responses to labour market issues are conceivable. In matters relating to economic redistribution, the different role assigned to the family may result in cross-case differences.

Results

The empirical results are reported in two blocks: Labour market issues and Social issues. Each section provides an account of identified problem definitions before moving on to policies and practices.

Labour market issues

All stakeholders mentioned workers’ rights, social dumping and tax evasion as implications of CEE citizens in informal work. They were perceived as getting lower wages, as being exposed to higher risk and as working longer hours than domestic workers (Online Survey AT 2014; Online Survey NL 2014; Online Survey SE 2014). A Swedish trade unionist describes the situation as:

They [the companies] know that they can pay these workers a lot less than native Swedes and thus earn a lot of money using underhand means. (Trade union representative 2014)

Stakeholders also said that informally working CEE citizens risked not getting paid at all (Online Survey AT 2014; Online Survey NL 2014; Online Survey SE 2014).

Turning to the practical phase, the identified problem definitions have resulted in various practices. Both Austria and the Netherlands initially implemented transitional rules, but current strategies vary.

An early policy measure in Austria was the Anti-Wage Dumping Law introduced in 2011 (ATAT-11-3 2015).⁸ The implementation of the law generated new networks, administrative structures and channels for knowledge transfer between social partners and public institutions (ATWI-3-3 2015).

Dutch efforts to uphold national labour market standards also resulted in legislation and programmes concerning wages, private employment agencies and increasing control (Letter to Parliament 2011; Ministry of Social Affairs and Employment 2013a, 2013b, 2014). A good example is the special certification of employment agencies, operated by the Ministry of Social Affairs and Employments (Letter to Parliament 2013). Dutch efforts were also concerned with access to data and registers. The Hague and Rotterdam have agreements with employment agencies to exchange information about registered workers to improve oversight (Municipality Rotterdam 2007, 2008b, 2011, 2012; Municipality the Hague 2010). In addition, Rotterdam actively tries to limit the number of temporary employment agencies (Municipality Rotterdam 2012, 2013b, 2014; Municipality Schiedam 2013, 2014).

Swedish responses to labour market implications include some legislative reform regarding posted EU labour, several government committees and information exchange platforms (Regeringen 2012, 2016a, 2016b).

In Austria and Sweden, the issues pertaining to workers’ rights tend to be considered the responsibility of trade unions. Their efforts predominantly consisted of various sorts of information points and help centres for undocumented workers (ATWI-12-3 2015; Online Survey SE 2014).

Policy responses towards the group of unregistered, non-educated CEE citizens varied. The Netherlands did not present any measures for this group, while a few initiatives were identified in Austria and Sweden. In Vienna, the city offers short-term employment to homeless CEE migrants as a step towards inclusion (ULLAT_15 2015). In Sweden, some NGO s provide this group with formal jobs by helping them register at the Employment Agency or hiring them themselves (SEGBG_15 2015; SEGBG_16 2015). Table 2 summarizes the policy responses in the labour market.

Table 2. Policy responses to labour market implications.

Phase 1.	Phase 2
Problem definitions	Policies and practices
	Austria	The Netherlands	Sweden
Social dumping Exploitation Lack of labour rights Tax evasion	Transitional provisions	Transitional provisions	Government committees and reports
	Social partner efforts	Fraud and illegality legislation	Social partner efforts
	Anti-Wage Dumping Law	Control initiatives	Venues for communication
	Assistance to destitute migrants (public)	Data exchange	Assistance to destitute migrants (NGO)

Social issues

Implications not related to the labour market include a wide range of issues. The issues are often interrelated and the subheadings should, therefore, be considered as structuring tools rather than rigid demarcations.

Non- registration

Stakeholders generally considered the fact that EU citizens without legal residence are “invisible in the records” as a problem. Often this is because of the unfortunate consequences for the individual. As explained by a Dutch stakeholder below, registration is a precondition for participating in society, e.g. paying taxes or accessing public services (Online Survey AT 2014; Online Survey NL 2014; Online Survey SE 2014).

So, they [CEE citizens] stay there in a vicious circle, because they can’t get legal work without a citizen service number [BSN]. But if they want a citizen service number, they need an address, which they don’t have because they can’t afford this. (NGO representative 2014)

Dutch stakeholders also perceived non-registration as a conscious strategy of migrants and employers to remain “under the radar”, as registration is sometimes perceived as disadvantageous:

They say: Oh, at the moment we register ourselves we get all kinds of letters, all kinds of taxes, and nothing remains. […] You also have people who do not (but) would like to register themselves because their boss tells them: ‘otherwise I don’t pay you anymore. (NGO representative 2014)

Policy responses reflect the dual function of registration as the gateway to services as well as to public audit and control.

EU citizens in Austria not applying for a residence certificate after three months may face a fine (ATWI-1-3 2015; Enengel and Reeger 2015). In order to enable homeless people to register, some NGOs in Vienna offer to serve as a main residence (ATWI_02 2014).

The Dutch context revealed several measures encouraging registration that aimed to facilitate control and improve official statistics. As explained by a local policy maker:

Registration is very important. […] With registration, a municipality knows how to steer […]. (Policy maker, Rotterdam, 2014)

Rotterdam and The Hague began to lobby for new registration procedures very soon after the Eastern accessions (Municipality Rotterdam 2008b, 2010, 2013b). The Ministry of Internal Affairs has responded with new legislation, e.g. facilitating clustering of public data (Ministry of Internal Affairs 2012, 2013). Some cities have improved existing registration practices, e.g. organizing group registrations outside office hours (Jansen and Tieleman 2015; Municipality Rotterdam 2013b; Municipality Westland 2013).

In Sweden, non-registration was primarily mentioned as decisive for local social services, when assessing benefit claims. The interest organization for local and regional authorities (SALAR) has issued guidelines for such assessments, as has the National Board of Health and Welfare (SALAR 2016; Swedish Board of Health and Welfare 2014). Since 2013, changed registration routines make it complicated to register residence for a stay shorter than one year (SE_07 2015). EU citizens may also register as job seekers at the Swedish Employment Agency (SESTH _02 2014; SESTH_14 2015; ULLSE_14 2014). Despite alarmist reports about arbitrary and inconsistent Agency practices in the processing of mobile EU citizens, few changes have been made. Local examples of informal cooperation between the agency and NGOs to facilitate the handling of cases involving informally residing CEE citizens were noteworthy exceptions (SEGBG__11 2015). Table 3 summarizes policy responses to non- registration.

Table 3. Policy responses to non-registration.

Phase 1	Phase 2
Problem definition	Policies and practices
	Austria	The Netherlands	Sweden
Non-registration	Information	Legislation (local and national level)	Legislation
	Counselling	Innovative registration practices	Informal cooperation (NGO - public sector)
		Data exchange Information dissemination	Guidelines

Informal housing

Informal work often coincides with informal accommodation arrangements. Respondents acknowledged how employers offering informal work with housing included risked creating multiple dependencies between employer and employee. One NGO representative said:

Those who have everything arranged are problematic as well, since, if their contracts end, they don’t have wages and are homeless. (Organisation for migrant workers 2014)

Austrian stakeholders reported that CEE citizens sometimes stay in private flats of inferior standard and in over-priced accommodation (ATWI_03 2014; ULLAT_14 2014). Housing was a dominant concern in the Dutch case, in particular the fact that many regulations are not applicable in cases of private housing arrangements (NLDH_03 2014). As a result, CEE migrants were perceived to be segregated in particular geographical areas and to be struggling with overcrowdedness (NLRD_10 2014; van Ostaijen et al. 2015). In the Swedish context, informal housing was mentioned but was not perceived as being specifically related to CEE migration (Online Survey SE 2014, SESTH_17 2013).

The different salience of housing across contexts is mirrored in the identified practices. Most measures were found in the Netherlands (Municipality the Hague 2008; Municipality Westland 2008). The Hague, followed by Rotterdam, managed the housing supply by a “two track policy”, aiming to provide affordable housing while simultaneously sanctioning illegal arrangements (Municipality Westland 2013). Rotterdam has also increased controls and sanctions overcrowding. After three fines, landlords of overcrowded premises lose their sublet permit (Municipality Rotterdam 2013a). The Hague runs a pilot project in which the control of housing arrangements is linked to residence registration (Municipality the Hague 2013b).

The other contexts did not reveal any specific policy responses to address informal housing. Temporary housing facilities in Linz may tentatively be a response to overcrowded housing, but was not explicitly framed as such (ATLI_12 2014). Table 4 provides an overview of policy responses relating to informal housing.

Table 4. Policy responses to informal housing.

Phase 1	Phase 2
Problem definitions	Policies and practices
	Austria	The Netherlands	Sweden
Informal housing preventing control of housing standards and rent levels. Crowded housing	Temporary housing provision by non- public stakeholders	Legislation Sanction systems Provision of affordable housing Data exchange (housing and registration)	–

Accompanying family members

Dutch and Swedish stakeholders mentioned accompanying family members as an implication of informal CEE migration. They mentioned how informal work often means unstable housing arrangements and landlord dependency, which makes it difficult for abused spouses, for example, to separate. Also, the absence of legal residence disqualifies these CEE citizens from regular systems of assistance (NLDH_09 2014; NLRD_12 2014; SESTH _18 2014). Dutch and Swedish respondents also mentioned the vulnerability of accompanying children (NLDH_01 2014):

I deal with parents who are working such long hours that it is impossible for us to speak to the parents. […] Or very sad cases where children fall asleep in classrooms because they awoke at four o’clock in the morning, at the time their parents have to work. (Education sector representative 2014)

The uncertainty of living as unregistered or homeless also made it difficult for accompanying children to perform in school (Online Survey SE 2014; SEGBG_09 2014; SESTH _06 2015).

No Dutch measures address the situation of family members. In Sweden, children without a proper home are considered as at risk of negligence by law and the social services have a duty to act. Accordingly, local administrative routines for such situations have developed, primarily informing the parent about the possibility of a paid return (SESTH _06 2015; SESTH_10 2015). Children of homeless EU citizens in Gothenburg may use an NGO-operated day care centre funded by the city (SEGBG_09 2014; SEGBG_16 2015). Gothenburg also grants the children access to education, while Stockholm advises their schools not to (SEGBG_10 2015; SESTH _06 2015). In cases of domestic abuse, one Swedish municipality reported “ad hoc provision” of emergency housing, despite lacking formal solutions (SESTH _18 2014). Table 5 shows the policy responses to accompanying family members.

Table 5. Policy responses related to accompanying family members.

Phase 1	Phase 2
Problem definitions	Policies and practices
	The Netherlands	Sweden
Abuse Accentuated dependency Children in risk environments	See efforts to reduce informal housing	Emergency assistance
		Paid return home
		Education or paid return home

Homelessness

Some EU citizens without legal residence arrive or become homeless in host Member States. As an Austrian stakeholder explained:

 …  being homeless in Austria is still more “charming” than being homeless elsewhere. (NGO representative 2014)

Austrian and Swedish respondents discussed the homelessness of CEE citizens to a greater extent than did the Dutch respondents. They also focused on the individual consequences rather than the societal impact of homelessness, which was at the centre of Dutch stakeholders’ concerns. They were unique in problematizing homeless CEE citizens in terms of a “crowding out” of the Dutch homeless (Municipality the Hague 2008). Rotterdam reported a low number of homeless CEE citizens overall (Municipality Rotterdam 2008a, 2008c).

Homeless CEE citizens were addressed in cross-sectoral policy endeavours, such as publicly funded NGO activities. In Vienna, NGOs provide most of the services available to unregistered EU citizens, co-funded by the city and/ or by private donations (ATWI-9-3 2015; ULLAT_15 2015; Vinzenzgemeinschaft Eggenberg- Vinziwerke 2015). Viennese stakeholders meet regularly to exchange information and to pass on insights to higher political and administrative levels (ATWI_02 2014). Linz handles homelessness differently, limiting the services available to day care centres run by NGOs or churches (ATLI_12 2014; ATLI-14-3 2015).

Dutch cities also restrict shelter accommodation available to EU citizens without legal residence (except in wintertime). Local government is allowed to require documented regional attachment before providing any services. The Hague reported monitoring homelessness prevalence as a preventative measure against “crowding out” effects (Municipality the Hague 2008; VNG 2011).

In Sweden, growing concerns among municipalities over homeless CEE citizens made the government appoint a national coordinator in 2014. The instruction was to encourage a coherent use of public authority in responses to destitute EU citizens and cooperation with countries of origin (SE_06 2015; SE_10 2015). Gothenburg and Stockholm operate outreach teams of social workers and provide emergency accommodation and basic services in cooperation with NGOs (City of Gothenburg 2013; City of Stockholm 2015; SEGBG_10 2015; SEGBG_16 2015; SESTH _06 2015; SESTH_10 2015). Table 6 summarizes policy responses to homelessness.

Table 6. Policy responses to homelessness.

Phase 1	Phase 2
Problem definition	Policies and practices
	Austria	The Netherlands	Sweden
Homelessness of CEE citizens	Cross-sectoral provision of emergency assistance. Multiple level information exchange	Monitoring Shelter (wintertime)	Cross-sectoral provision of emergency assistance.
			National coordinator

Begging

Problem definitions relating to begging are varied. Austrian and Swedish stakeholders identified begging as an issue related to intra-EU mobility that is causing heated public and political reactions (ATLI-14-3 2015; ATLI-15-3 2015; SE_06 2015; SEGBG_09 2014). In the Netherlands, beggars were a rare sight and not considered an implication of CEE migration (NLRD_12 2014; NLRD_14 2014).

Practical measures ranged from criminalization to lobbying efforts. Rotterdam and The Hague forbid begging (Municipality Rotterdam 2008a). Linz and Vienna criminalize certain forms of begging, such as commercial begging or begging with children (Aichinger and Winroither 2013; Verfassungsgerichtshof 2012; Winter 2015). Unique to Austria were organizations in Vienna and Linz opposing the ban, advocating de-criminalization and providing legal advice to beggars (ATLI-15-3 2015). In Sweden, actively approaching someone for money theoretically requires a permit. More passive forms of begging are allowed (Swedish National Police Board 2014). In addition to general outreach efforts, a “hotline” for concerns relating to destitute EU citizens in Stockholm was the only practice relating to begging. Table 7 summarizes the policy responses to begging.

Table 7. Policy responses to begging.

Phase 1	Phase 2
Problem definition	Policies and practices
	Austria	The Netherlands	Sweden
Begging in public space	Regulation Private- Public enforcement Lobbying for recognition Information point	Prohibition	“Hot line”

Stranded CEE citizens

Vienna, Gothenburg and Stockholm respondents faced “stranded” CEE migrants, that is, persons that lack a right to residence as well as the resources to return home (ATWI_02 2014; SEGBG_30 2013; SESTH_40 2014; SEGBG_10 2015; SEGBG__07 2014; ULLSE_14 2014). Dutch stakeholders only recognized the problem implicitly when discussing homelessness.

All cities had routines for encouraging or enforcing the return home of destitute CEE citizens. In Vienna, NGOs assist people with travel arrangements and starting over at home (ATWI-9-3 2015; Caritas 2015). The Hague and Rotterdam have invented models for voluntary return in cooperation with a Polish-Dutch NGO, now operating on a national level. In addition, The Hague addresses “stranded” CEE citizens with a threefold approach. The first track consists of voluntarily returns executed by NGOs. The second track, operated at the national level, is activated when a CEE citizen exerts “a disproportionate pressure on the Dutch social system”. The third option is “coercive return”, used if someone presents “a serious threat for fundamental parts of Dutch society” (Municipality the Hague 2013a). Rotterdam runs a contact point for stranded EU citizens and hosts a national initiative relocating CEE citizens with a criminal record (Municipality Rotterdam 2008a, 2010, 2013a).

Swedish municipalities reported assisted returns to be the most common support offered to homeless EU citizens in 2014 (SALAR 2014). The city of Stockholm has occasionally offered paid returns without the individual assessment normally required. Lately, as the number of destitute CEE citizens has increased, Stockholm and Gothenburg have become more reluctant to offer paid returns (City of Gothenburg 2013; SEGBG_30; SESTH_25 2013; SESTH _06 2015, SESTH_40 2014). Table 8 summarizes the policy responses to stranded CEE citizens.

Table 8. Policy responses to “strandedness”.

Phase 1	Phase 2
Problem definition	Policies and practices
Destitute CEE citizens without resources to return (“Strandedness”)	Austria	The Netherlands	Sweden
	Paid voluntary returns
		Forced repatriation in cases of lacking right to residence Forced repatriation in cases of criminal conviction	Repatriation in cases of criminal conviction

Health care access

Access to health care for CEE citizens in informality was problematized in relation to their general situation of vulnerability (ATWI-9-3 2015; Letter to Parliament 2011; SESTH_09 2015). Vienna and Stockholm stakeholders specifically mentioned mental illness among homeless CEE citizens as being a concern:

This group is a big mystery to the social services. We don’t know how they manage or where they come from. (City administration 2013)

In terms of practices, most initiatives providing health care for informal CEE citizens operate locally. The nature of arrangements ranged from permanent and formal to ad hoc and informal.

In Austria, the main providers of healthcare to this group are NGOs and confessional hospitals. In addition, the city of Vienna offers short-term convalescence accommodation, and non-public actors in Vienna and Linz operate mobile medical offices for the homeless (ATWI_02 2014; ATWI_09 2014; ATWI-9-3 2015; Caritas der diözese Linz 2015).

Dutch measures are limited to small, local initiatives, such as free-of-charge nurses (Letter to Parliament 2011; Ministry of Spatial Planning 2008; Municipality Rotterdam 2008a).

In Sweden, access to health care for EU citizens without legal residence is dominated by informal arrangements. Although there are national guidelines for public health care, counties follow different routines for billing and treatment (SEGBG_14 2015; SEGBG__11 2015; SESTH_09 2015; Swedish Board of Health and Welfare 2015). A respondent describes the situation below:

… this world [of CEE citizens in informality] constitutes a kind of shadow society. Either there are informal agreements, or none. […] Because people don’t know the guidelines, or the established practice, what you are supposed to do, but instead have some made-up-idea, also among public authorities and health care organisations, as I understand it. (NGO representative 2015)

In Stockholm, CEE citizens without legal residence get health care at an NGO-operated clinic. In Gothenburg, the same group is referred to a publicly funded centre (SEGBG_12 2015; SESTH_09 2015). Several national and European projects also address the issue of the health care access of undocumented persons in general (City of Gothenburg 2015; Karolinska University Hospital 2015). Table 9 summarizes policy responses to the health care needs of CEE citizens without legal residence.

Table 9. Policy responses to health needs.

Phase 1	Phase 2
Problem definition	Policies and practices
Physical and mental health issues of unregistered EU citizens	Austria	The Netherlands	Sweden
	Cross sector health care provision Private-NGO health care provision	Local initiatives	Cross sector networks Public health care provision NGO health care provision

Analysis

Figure 4 summarizes the findings by categorizing policy responses as one of the theorized outcome alternatives.

Figure 4. Policy responses to informal CEE migration per implication.

The figure only contains complete policy processes. Cases in which problem definitions were not followed by accompanying practices are left out of the analysis.⁹

The results show that some policy responses unambiguously focus on either need or informality and some shift focus during the policy process. Labour market implications were handled similarly across contexts, while policy responses to social issues varied, in particular between the Netherlands and the other cases, Austria and Sweden.

Policy responses relating to the labour market were expected to emphasize informality. This was confirmed, with shifting emphasis across contexts. As shown in Figure 4, stakeholders invoked both need and informality when defining problems related to the labour market. Actual policies and practices targeting such issues however often focused on sanctioning informality. The Austrian and Swedish context considered workers’ rights as a matter of need, but concrete measures were delegated to trade unions and civil society.

Responses to social issues were mixed, with more similarities between Austria and Sweden than between either of them and the Netherlands. In cases of extreme need, such as visible begging, homelessness and strandedness, Swedish and Austrian policy responses signalled a commitment to addressing need. Only Sweden problematized and addressed the situation of accompanying family members as a matter of need only. The priority given to addressing need in the social domain was expected, at least in the most tangible instances of need. This was the case in Sweden and Austria, while Dutch responses instead focused on informality, responding to homelessness, non-registration, strandedness, informal housing and begging.

Conclusion

This article makes empirical, methodological and theoretical contributions to the state of the art. The results show similar responses to implications relating to the labour market. The nuances of corporatism did not result in dramatically different policy responses. There was a larger discrepancy in policy responses to social issues.

The conservative and universal welfare regimes problematized and addressed social implications as matters of need as well as informality. The exception was homelessness, being unambiguously treated as a matter of need in both Austria and Sweden. The formality dimension was secondary to the needs dimension, indicating a solidarity based response. None of the Dutch policy responses related to need only, but instead addressed homelessness, begging and strandedness as matters of informality. Formality was given priority over need, indicating that solidarity centred explanations seem less valid in the Dutch regime.

Results indicate that conservative and universal welfare states are equally sensitive to distributive outcomes and procedural justice. The fact that only Swedish responses dealt with accompanying family members indicates that the different role of the family in service provision affects policy. The results also seem to indicate that the institutional contexts of Austria and Sweden allow for alternating between addressing need and sanctioning informality, at least when responding to social issues. The Dutch mixed regime instead prioritizes procedure by consistently sanctioning informality. The findings are helpful to analysing other Member States, e.g. France, Germany and Denmark, witnessing similar debates about social dumping and welfare tourism.

Methodologically, the conceptualization of the policy process proved useful in cross-case analyses of policy responses. Distinguishing between need and informality centred policy responses allowed for a deeper understanding of the underlying mechanisms in encompassing regimes. The focus on problem definitions and practices opens up for questions about the consequences of policy responses. Future research exploring the effect for individual migrants and receiving societies are most welcome.

Acknowledgements

The author is grateful to the anonymous reviewers for constructive comments on earlier versions of the paper.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributor

Karin Zelano is a PhD Student at the Department of Political Science since 2013. Research areas of special interest include migration, intra-EU mobility, informal work, welfare state institutions and governance.

ORCID

Karin Zelano http://orcid.org/0000-0002-4219-5913

Additional information

Funding

This work was supported by the Joint Program Initiative Urban Europe under Grant number 438-12-412 (IMAGINATION).

Notes

1 Case C-140/12, Peter Brey [2013].

2 Case C-333/13, Elisabeta Dano and Florin Dano [2014].

3 Case C-67/14, Nazifa Alimanovic and others [2015].

4 Presented for the first time in Case C-85/96, Maria Martinez Sala [1998] ECR-I-2691.

5 Austria: Vienna, Linz. The Netherlands: The Hague and Rotterdam. Sweden: Stockholm and Gothenburg.

6 A full list of interview respondents is available in the Appendix.

7 Country-specific problem definitions included: Working conditions in 24-hour care work (AT), Risk of poverty retirement (AT), EU citizens as victims of crime (SE), Geographical clustering (NL).

8 Amended in 2015.

9 Vulnerability of accompanying family members (NL), Informal housing (SE).

Appendix: cited interview respondents

Table

Abbreviation	Respondent	Interviewer(s)
Austria
National level
ATAT-11-3. (2015)
	Federal Ministry for Europe, Integration and Foreign Affairs	Reeger, U. and Enengel, M-L.
Local level
Linz
ATLI_12. (2014)	Integration Office, Traun
ATLI-14-3.(2015)	migrare – Zentrum für MigrantInnen Oberösterreich (Advisory centre for migrants), Linz
ATLI-15-3. (2015)	Caritas Linz
Vienna
ATWI_02. (2014)	Day-Care Centre “Zweite Gruft”, Caritas Vienna
ATWI_03. (2014)	Helping Hands, Vienna
ATWI_09. (2014)	Sozial- und Rückkehrberatung, Caritas Vienna
ATWI-1-3. (2015)	MA17 (Integration and Diversity), City of Vienna
ATWI-3-3. (2015)	WAFF (Vienna Employment Promotion Fund)
ATWI-4-3. (2015)	Helping Hands, Vienn
ATWI-9-3. (2015)	Day-Care Centre “Zweite Gruft”, Caritas Vienna
ATWI-12-3. (2015)	Chamber of Labour, City of Vienna
The Netherlands
Local level		Van Ostaijen, M.
The Hague
NLDH_01. (2014)	NGO representative
NLDH_02. (2014)	NGO representative
NLDH_03. (2014)	Street-level bureaucrat, police officer
NLDH_09. (2014)	Temporary employment agency
Rotterdam
NLRD_10. (2014)	Language policy representative
NLRD_12. (2014)	Street-level bureaucrat
NLRD_14. (2014)	Policy maker
NLRD_15. (2014)	Education sector
Sweden
National level
SE_03. (2014)	Legal expert at The Swedish Tax Agency	Zelano, K.
SE_06. (2015)	Swedish Association of Local Authorities and Regions (SALAR)
SE_07. (2015)	Swedish Tax Agency
SE_10 (2015)	Interview with national coordinator
Local level
Gothenburg
SEGBG_01. (2014)	The Building Workers Union, Gothenburg
SEGBG_09. (2014)	Göteborgs Räddningsmission (NGO)
SEGBG_10. (2015)	Social services, City of Gothenburg
SEGBG_12. (2015)	Care centre for the Homeless (Public)
SEGBG_14. (2015)	Red Cross GBG (NGO)
SEGBG_15. (2015)	Faktum (Street paper) (Private)
SEGBG_16. (2015)	Göteborgs Räddningsmission (NGO)
SEGBG__03 (2014)	The Swedish Public Employment Service (Gothenburg City)
SEGBG__07. (2014)	Outreach team, Social welfare office, City of Gothenburg
SEGBG__11. (2015)	Crossroads GBG (NGO)
SEGBG_30 (2013)	City of Gothenburg, Unit for social resources
Stockholm
SESTH _02. (2014)	The Swedish Public Employment Service (Stockholm City)
SESTH _05. (2013)	Mikamottagningen, City of Stockholm
SESTH _06. (2015)	Coordinator Vulnerable EU citizens, Social services, City of Stockholm
SESTH _18. (2014)	Municipality of Haninge, the unit for economic support
SESTH_09. (2015)	Läkare i Världen (NGO)
SESTH_10. (2015)	Ny Gemenskap (NGO)
SESTH_14 (2015)	Personal communication [Forskning: Mobila EU-medborgare].
SESTH_25. (2013)	The unit for social services, City of Stockholm
Focus groups
ULLAT_14. (2014)	Vienna Urban Living Lab: Urban Implications.
ULLAT_15. (2015)	Vienna Urban Living Lab: Governance,
ULLSE_14. (2014,	Gothenburg Urban Living Lab: Urban Implications.