Department of Political Studies - University of Catania
Jean Monnet Chair of European Comparative Politics

Jean Monnet Working Paper in Comparative and International Politics


European Forum Fellow, EUI, Florence

Co-Director, Mediterranean Migration Observatory, Athens






September 1997 - JMWP 12.97


The relationship of non-citizens to European welfare systems is perhaps one of the most complex yet under-researched areas of contemporary policy. What little comparative literature exists spans the whole gamut of possible analyses of this relationship: from Freeman (1986) who suggests that European welfare states are almost uniformly exclusive owing to their citizen-based provision; through Baldwin-Edwards (1991) who argues that different states interact very differently with their immigrant populations; and more recently Soysal (1994) and Jacobson (1996) who both claim that European countries are almost uniformly inclusive of their resident aliens owing to the massive impact of international human rights law.

 At the end of the twentieth century we have the embarrassment that not only is there no consensus on the actuality, but also the issues are becoming politically dangerous. Increasingly, immigration is blamed for the financial crisis in sustaining generous welfare arrangements, particularly with the highly visible assistance given to asylum seekers and refugees (Bauboeck, 1994). Almost all available evidence suggests that immigrants contribute far more than they receive from welfare arrangements (e.g. Barabas et al, 1992; Ruland, 1994). Another related claim is that immigration causes unemployment; yet both legal and undocumented migrants work in highly segmented labour markets with little or no competition from the indigenous labour force. Significantly, as immigrants have a high concentration in vulnerable industries such as textiles, there has been rising unemployment amongst the immigrant populations (Bauboeck, 1994). This is sometimes taken as clear evidence that immigration itself causes unemployment.

 As almost all European countries approach their predicted demographic crisis, with its problem of many pensioners to be supported by few of working age, it may seem that immigration would be an ideal solution. There are three major problems with this, however. First, the level of immigration would have to be phenomenally high - and this would be unacceptable politically. Secondly, inevitably these ‘guestworkers’ too would remain - thus changing the ethnic composition substantially after some time. Thirdly, the pension schemes of at least Germany have been sustained for some time by migrant workers’ contributions: Ruland (1994: 85) comments that for 1989 immigrants’ contributions exceeded payments by a factor of three or four, thus bringing short term relief to the system.

 Finally, it is worth noting that whereas expanding European economies thirty years ago were able to absorb immigrants and also develop extensive welfare systems, today unskilled labour is mostly unwelcome whilst those welfare systems are undergoing major reform. The underlying forces are, of course, economic; yet the question of policy options remains open and largely ignored with respect to migrants.

 In this paper I shall sketch the patterns of evolution of European welfare states, their various phases and the broad relations with immigrants. Then I address the question of the welfare needs of migrants, by migrant category and also using life-cycle hypotheses; the relative successes of different welfare systems in satisfying these needs are then examined. Finally, some policy options are proposed as plausible directions for the future.



 Welfare states emerged across Europe at different times, with different programmes and with different institutional forms (Esping Andersen, 1990). Strong and ethnically homogeneous labour movements were one possible common factor in the evolution of European welfare states (Stephens, 1979); it has been argued (Faist, 1995) that ethnic heterogeneity - such as found in the USA - led to a small welfare state and race-class cleavages.

 There can be little doubt that welfare systems are explicitly tied in with the evolution of most nation states and their national citizenship (Marshall, 1950); further, that it is unclear how immigrant populations should relate to the welfare system. The available evidence suggests that as immigrant populations were not formative in their development, their participation was not considered (Crowley, 1998). Formal participation could be enabled by acquisition of legal nationality: other intermediate statuses were perhaps to be negotiated.

 Following Esping Andersen, we can identify four welfare regimes in Europe: the social democratic [Scandinavia]; conservative [continental Europe]; liberal/social democratic [UK[; and the southern European (Esping Andersen, 1990; Baldwin-Edwards and Gough, 1991; Ferrera, 1996). The particular arrangements of these regimes have implications for immigrant participation, as we shall see below.

 The social democratic model of welfare has as its linchpin the concept of social citizenship, that is to say a universalist approach, tax-financed, residence-based and delivering high levels of benefits with top-up schemes maintaining some differentiation. Necessarily, this system is expensive, and requires high levels of employment and taxation.


 The UK system is a mix of liberal/social democratic, with the former predominating in the last two decades. Rights to social benefits in law are weak; financing is a mix of contribution and taxation; only medical services via a distinct National Health Service are universal, otherwise benefits tend to be means-tested; occupational and status differentiation are low. The private sector is extensive and runs alongside a centralized state sector.

 The conservative regime exhibits law-based rights to a wide range of benefits, with social insurance as the principal mode of organization and delivery of services, with occupationally differentiated benefits delivered by para-state institutions. There exists also a set of means-tested benefits for those who fall outside of the insurance system.

 The southern European model resembles an under-developed version of the conservative regime. However, there exist some distinct differences which encourage a separate categorization. First, the coverage of the population is low, with very low or no benefits means-tested for those who fall outside. Secondly, the differentiation of benefits is very high - far greater than would reflect occupational earnings differentials, with privileged groups obviously benefiting. Thirdly, a massive

asymmetry of pensions expenditure alongside underdeveloped unemployment benefits and inadequate universalistic national health systems. Finally, the management of the para-state funds is non-transparent and in certain cases heavily subsidized by taxation in clientelistic fashion (Ferrera, 1996).


The incorporation of immigrants into welfare systems

 There are, I suggest, four basic mechanisms by which immigrants can be incorporated into a national welfare system:

  1. by ‘structural acceptance’ or toleration of the system
  2. through ‘denizenship’ or some explicit active immigrant policy
  3. by treaty for ‘privileged aliens’
  4. through international human rights norms enforced by the courts i.e. case law


Throughout the postwar period, an encyclopaedic collection of multilateral and bilateral treaties has emerged, in order to cope with the reality of increasing migration of workers and also to protect generally human rights (Hollifield, 1992; Baldwin-Edwards, 1991).

 Table 1 shows some of the bilateral treaties signed for labour recruitment. They all grant social security equality with nationals; a few allow for vocational training; very few grant reciprocal coverage.


Table 1

Labour Recruitment Agreements








Nationals of:




























































 SOURCE: Baldwin-Edwards (1991: 195)

Note: * indicates treaty date unknown


Other multilateral conventions governing migrants’ social rights include the European Convention on Human Rights and additional Protocols; the European Convention on Establishment, 1955; the European Social Charter, 1961; the European Convention on Social Security, 1972. Of these, the most significant for case law is the ECHR; however, many EU countries have signed the other conventions and incorporated them into domestic law.

 The extensive labour migration of this period coincides with the expansion and consolidation of most European welfare systems. Yet the patterns of public policy, although showing some degree of convergence, still retain clear differences (Esping Andersen, 1990). What seems uniquivocal is that migration has had little structural impact on welfare systems. The modifications have been piecemeal and fragmented attempts to rationalize what are increasingly irrational nationally-focused welfare systems in an interdependent global economy.

 Using the typology of welfare regime, we might expect the social democratic to be the most inclusive, the conservative to benefit migrants after substantial periods of employment and residence, the UK to be ambiguous, and the southern European to be exclusive. In broad terms this analysis holds, although there are specific problems associated with each regime type.

 Only two European countries [Sweden and Netherlands] have developed active immigrant policies (Hammar, 1985). Others have granted equality with nationals for their recruited guestworkers in certain areas of social policy; most have been obliged to concede substantial social rights for their long-term residents, often referred to as ‘denizens’. However, the granting of permanent residence rights varies greatly across Europe, as shown in Table 2. Only recently has permanent status become a (theoretical) possibility in southern Europe; elsewhere, the qualifying periods of residence are highly divergent.


Table 2

Residence period required for permanent status






















 SOURCE: Niessen (1989) cited in Soysal (1994: 121)


Interestingly, these differing requirements of aliens law [immigration regimes] coincide almost precisely with the Esping Andersen typology of welfare regimes (Baldwin-Edwards, 1991; Faist, 1995). France is anomalous, although recently has shifted to a position more typically continental. The southern European countries’ underdeveloped immigration systems correspond well with the level of development of their welfare systems, with Greece and Portugal lagging well behind in both areas.

 Finally, it should be understood that it is not possible to identify migrants’ social rights without analyzing their relationship to aliens law. Theoretical entitlements can be misleading, since claims can lead to refusal of residence permit renewal. The strongest claims are by denizens and those with Convention refugee status. Others may risk their residence status in claiming social rights.

 One perverse "negative welfare" effect is the general requirement of a minimum income and suitable housing for a migrant to exercise the right to family reunion. This minimizes the potential of welfare claims by migrants, despite the fact that they may have contributed fully in insurance and tax contributions. Across the EU, only two states do not have this requirement - Denmark and Sweden.




 The greatest problem in tackling the analysis of the relationship of immigrants and welfare systems is the complexity of the matter. In each country, two things are critical for this relationship: the legal status of the migrant and the migrant’s nationality. The first is important because of the interaction between aliens law and social law; the second because of the existence of bilateral or multilateral treaties benefiting specific nationals. Of specific note are EU arrangements: other than the well-known rights of EU migrants, there are two other mechanisms through which third country nationals may benefit. The first is simply as a family member of an EU national; the second is through the direct applicability of agreements entered into by the EU and a third party. In particular, of note are the EU/Turkey Assocation Agreement and related Decisions; the EU/Maghreb Co-operation Agreements. These give substantial rights to legally resident nationals in the EU, usually approximating to nationals’ and EU citizens’ rights.

 Various legal categories exist; for our purposes we can identify the following:


A sixth and major category is that of asylum-seekers. The situation of these is constantly changing and impossible to assess. Generally, they are prevented from working and allotted assistance benefits at a lower rate than the indigenous population. There is a particular problem in southern Europe where few facilities are provided; the worst provision appears to exist in Greece.

 Table 3 gives a very rough guide to the situation of migrants by category. This table ignores the provisions benefiting ‘privileged aliens’, by treaty or EU law. Necessarily, there are variations in the treatment of these groups across Europe: the classification given indicates the general position only, with great variation indicated by "?".



Table 3

 Characterization of immigrants’ welfare rights





Unemp. ins.

Social assist.

Public housing













perman. residents






Convention refugees






tolerated refugees







Y = generally available

? = wide variation and/or doubt

N = generally not available


 The welfare service most readily given is school education for migrants’ children, although access to higher education is much more restricted. Healthcare also is generally available - even to illegal migrants. Unemployment benefits are available if the migrant satisfies the conditions of the scheme; social assistance [means-tested] is rarely available and can threaten the residence status of the applicant. Public housing is sometimes available in principle, but in practice immigrants are given low priority.

 Looking at benefits by category of migrant, the best-protected are Convention status refugees and those with permanant residence rights [denizens]. The least protected are illegal immigrants and workers, followed by ‘tolerated’ refugees.


The Social Needs of Migrants

 Little attention has been paid in the literature to the life cycle of immigrant populations. Since the principal effect of most social policy regimes is to redistribute in the life cycle, rather than across the general population, this might reflect the general lack of interest and research in the area. Furthermore, unlike the life cycle of indigenous populations where the greatest burden is found in youth and old age, the immigrant life cycle seems to need most support in its intermediate stage. This should have implications for social policy.

 We can characterize the ‘typical’ immigrant life cycle as having three phases:

 1] Guestworker. In this phase the existing legal provisions of many bilateral recruitment agreements may be adequate. The worker is likely to make few demands on the welfare services, owing to his/her age, and be a net contributor to the welfare system. Nevertheless, adequate healthcare should be provided along with access to decent housing.

 2] Family reunion. When joined by his/her family, the needs of the migrant are far greater. Obviously families with children are likely to need most: education, healthcare, housing. Also social protection should extend to involuntary unemployment schemes - including social assistance.

 3] Retirement. At the end of working life, many immigrants choose to be repatriated. It is imperative that they should have the financial resources to permit this, which basically means the right to ‘export’ the pension to which they have entitlement. This requires either the general possibility of exportable benefits in the welfare system, or specific provision by treaty for certain nationalities.

 This life cycle hypothesis, however, characterizes the traditional guestworker migrant. The principal forms over the last ten years have been family reunion, illegal migration and asylum-seeking. The last two present serious difficulties - for both the state and the migrant.


Patterns of Policy

 Phase 1

The available information in this area is very little. Given that few guestworkers are being recruited (except for highly skilled professional work) the first phase of the life cycle seems to apply principally to illegal migrants and undocumented workers. Clearly the level of protection is very low across the EU, although we might expect one country - the UK - to come out looking more generous. This is because the UK welfare system is to a large extent residential and tax-based, with the assumption of legal residence of all in the territory. (This assumption exists because of the emphasis on external border control in immigration policy.) What little evidence exists tends to support this hypothesis.

 The social democratic regimes seem to be less tolerant of illegal migrants, owing to the high degree of regulation in the system. There is probably little problem as illegal migration is better detected.

 Generally, there is a problem with illegal residents in continental Europe. In the conservative regimes, they are likely to exist completely outside of the system, whereas in southern Europe limited evidence suggests that insurance contributions are exacted from employees who will never be able to claim the associated benefits.


 Phase 2

The recent research of Maydell (1995) across 17 states concludes that there is little discrimination against legally resident third country nationals in social law, although there is a lack of transparency in many countries. They confine their Report to the social provisions of EEC Reg. 1408/71 [the co-ordinating rules for EU nationals] thereby omitting two problem areas of social assistance and housing; furthermore, the study omits any consideration of interaction with aliens law. These is no consideration given to empirical research, and great emphasis is placed on theoretical legal provision.

 It is generally perceived that the social democratic regimes present few problems, with substantial support for long term residents and their families. In other countries, the interaction with aliens law is crucial: if the head of household has permanent residence status, there may be few problems other than poor housing and potential unemployment. Without this status, it may even be impossible to achieve family reunion. One perverse effect noted by Bauboeck (1994), is that family reunion may be effected illegally owing to the difficulty of achieving it legally. Subsequently, children are not schooled in order to avoid detection of the family. In southern Europe, this phase is completely neglected. Not only are the welfare systems highly skewed, but the aliens regulations award few permanent residence. This, in combination with the very large number of illegal aliens, represents the most socially unprotected geographical area for migrants.

 Phase 3

Recent research has started to shed some light on this area. Bolderson and Gains (1994), in their study of exportability of benefits in the OECD, link the outcomes with the original structures of the welfare systems. The insurance based systems might be expected to grant exportability of pensions and some other benefits: in fact, the conservative regimes (in their study, France and Germany) in principle prohibit such claims. Two southern European countries - Italy and Portugal - allow exportability as a principle. One other country, the Netherlands, is noted as constructing its pension system deliberately to allow exportability, in contrast with its other benefits. The Swedish system has serious difficulties with restrictions and - like France and Germany - discrimination against non-nationals in its arrangements. The UK is seen as discouraging portability of pensions in principle, and it makes no distinction on nationality grounds. Generally, they conclude, the most difficult regimes for exportability are non-contributory and means-tested.

 The generalized ‘structural’ study cited above does not account for the existence of bi- and multi-lateral treaties derogating from the general principles. Ruland (1994) examines such provisions in the case of Germany, concluding that recent changes will improve the situation by requiring the exportability of all insurance benefits, although without treaty provision the estimated value of the Federal subsidy will be deducted. Only 56% of insured foreigners are covered by explicit treaty. The data supplied of payments made are surprisingly low, especially for Turks who are covered by both bilateral treaty and EU law. Thus a possible impediment exists, such as bureaucratic difficulty or simply ignorance of legal entitlement.

 Bonniol (1994) examines the 15 African treaty arrangements in the case of France. She offers few concrete conclusions, except to note that pensions provisions are highly developed and that there is great complexity and difficulty of implementation with such a variety of different schemes. For 1990, 86,000 families received family allowances and 200,000 pensions and annuity payments were made.

 Tamagno (1994) notes that there are few agreements made with developing countries - most are between developed countries. He attributes this to lack of political pressure from immigrant communities in developed countries, along with a low priority accorded by the governments of developing countries.

 Our conclusion must be that there is an absurdity in the structures of welfare systems which have adapted ad hoc by multiple treaty arrangements. Furthermore, the coverage is not good. It may be that the bureaucratic cost of implementing these complex arrangements is at least as great as simply conceding the general principle of exportable contributory benefits.



 We have seen that the social protection of migrants across the EU is fragmented and sometimes tenuous: not only is the variation between countries very great (as we would expect) but the differential protection afforded in different phases of the immigrant cycle, between different legal statuses and between different nationalities is also very great. Thus the claims of Soysal and Jacobson are exaggerated and unhelpful in assessing policy outcomes and options. On the other hand, Freeman is quite incorrect in asserting that European welfare states exclude immigrant participation.

 The piecemeal adaptation of welfare systems to immigration and the needs of migrants has been ad hoc, juridical, and unnecessarily costly and difficult to implement. Since welfare regimes are now undergoing (or about to undergo) radical restructuring there is the possibility of incorporating non-citizens in a rational ordered manner. The OECD suggests in a recent report that current labour market trends should be the basis of social policy (OECD, 1996: 21); that income transfers should be used less and other measures found to help the social and labour market position of those in need. An interesting point is made about the role of the non-profit (or NGO) sector, which is identified as important in delivery of welfare services. This is supported by the research of Room and 6 (1994) who mount a critique of Esping Andersen’s class coalition based approach, in that it ignores mobilization in the non-governmental sector. Although NGO provisions predominate in the less-developed welfare systems of southern Europe (with the exception of Greece), and play a significant role in ‘conservative’ regimes, this sectoral location of provision is likely to be a poor predictor of behaviour. Furthermore, as Room and 6 show, whereas the third sector traditionally has been subsidized by the state, increasingly NGOs draw on EU resources and even engage in lobbying at Community level (Room and 6, 1994: 58-60).

 It appears that the role of the EU in developing social protection of migrants’ rights is crucial. Already the jurisprudence of the Court of Justice has benefited considerably those nationalities covered by third country agreements. Suggestions of extending the EU co-ordinating rules on social security (Reg 1408/71) to third country nationals are now being made (e.g. Pieters, 1995); in fact, the Court has more or less taken as a baseline these provisions in interpreting the Assocation and Co-operation Agreements. This extension to all legally resident third country nationals would eliminate many of the problems and injustices in Phase 3 of the migrant life cycle.

 What of activity at the national level? The trend of governments has been to diminish the rights of (legal) migrants, whereas courts have been enforcing the established rights and even extending them. Judicial activism is now one of the principal innovating forces in this area. Several recent cases demonstrate this: in Germany, in Molenaar, Case C 160/96, the Court found that the non-exportability of pensions was unconstitutional . Also in Germany, the Administrative Court of Gottingen found that a recognized refugee and his family - expelled for claiming social assistance - must be allowed to remain in order that the family remain united (Az: 1B 1260/96). The European Court of Human Rights, in a case brought against Austria, found for the plaintiff - a Turk who had applied for social assistance when his unemployment benefit ran out. The ECHR found unlawful discrimination in property rights under Article 1 of Protocol 1 of the ECHR. Similarly, the ruling in February 1997 of the Court of Appeal in the UK, where local authorities’ appeal against the application of the National Assistance Act of 1948 in allowing asylum seekers to claim housing assistance and social security benefits, must be seen as a landmark decision. Essentially, the UK’s Asylum and Immigration Act, 1996 removed the right of an estimated 13,000 asylum seekers to claim benefits; the courts restored them, in what amounts to constitutional fashion.

 On the other hand, states continue their denial of immigrants’ social security claims. Austria, in June 1996, cancelled all bilateral social security agreements with Yugoslavia, Tunisia and Turkey - on the grounds that they were too expensive. There are expected to be cases brought before the European Court of Justice, particularly involving the Turkey Assocation Agreement. Even in Denmark , the trend is against migrants. Those with temporary residence permits who claim social assistance can now be denounced by local authorities: in 1995, 150 were deported in this fashion (MNS, 5/96).

 What of the future? The mobilization of migrants, particularly through migrant associations at both national and European level, must be a major possible mechanism for reform. Non-governmental organizations [NGOs] can be not only a socio-political force, but also have the unique capacity to incorporate and represent illegal migrants. Furthermore, the migrant groups may well constitute an unattached electoral force in a future ‘federalized’ Europe: such political power - along with NGO self-help, albeit with state and EU financial support - constitutes the most likely route for the emancipation of Europe’s migrant communities.




 Baldwin-Edwards, M (1991): ‘The socio-political rights of migrants in the EC’, in G Room (ed): Towards a European Welfare State?, Bristol: SAUS pp 189-234

 Baldwin-Edwards, M and I Gough (1991):’European Community social policy and the UK’, in N Manning (ed): Social Policy Review 1990-91, Harlow: Longman pp 147-168

 Barabas et al (1991): ‘Gesamwirtschafteliche Effekte der Zuwanderung 1988-91’, RWI-Mettelungen, vol 4 no 2, pp 133-154, Rheinisch-Westfälisches Institut für Wirtschaftsforschung

 Bauboeck, R (1994): ‘The Integration of Immigrants’, mimeo, Report prepared for the Council of Europe

 Bolderson, H and F Gains (1994): ‘Comparison of arrangements for exporting benefits relating to age, disability and widowhood in twelve OECD countries’, in ISSA (1994) pp 53-74

 Bonniol, F (1994): ‘Coordination of social security legislation between developed and developing countries: The relations between France and fifteen African countries’, in ISSA (1994) pp 211-226

 Crowley, J (1998): ‘The National Dimension of Citizenship in T. H. Marshall’, in M Baldwin-Edwards, R Bauboeck and G Zincone (eds): Whose Europe? Immigration and Citizenship into the 21st Century, Basingstoke: Macmillan [forthcoming]

 DRICSS (1995): Social Security in Europe: Equality between Nationals and Non-nationals, Oporto: Departamento de Relações Internacionais e Convenções de Segurança Social, [Proceedings of a Conference held in Oporto, November 1994]

 Esping Andersen, G (1990): The Three Worlds of Welfare Capitalism, Cambridge: Polity Press

 Faist, T (1995): ‘Ethnicisation and Racialisation of Welfare State Policies in Germany and the USA’, Ethnic and Racial Studies, 18

 Ferrera, M (1996): ‘The Southern Model of Welfare in Social Europe’, Journal of European Social Policy, 6/1, pp 17-37

 Freeman, G (1986): ‘Migration and the political economy of the welfare state’, AAPSS 485, pp 51-63

 Hammar, T (1985)(ed): European Immigration Policy, Cambridge: CUP

 Hollifield, J (1992): Immigrants, Markets and States, Cambridge: Harvard UP

 ISSA (1994): Migration: a Worldwide Challenge for Social Security, Geneva: International Social Security Association, Studies and Research No. 35

 Jacobson, D (1996): Rights Across Borders: Immigration and the Decline of Citizenship, Baltimore: Johns Hopkins Press

 MNS (as dated): Migration News Sheets, Brussels: Churches Coumcil for Migrants in Europe

 Marshall, T H (1950): ‘Citizenship and social class’, in T H Marshall and T Bottomore (1992): Citizenship and Social Class, London: Pluto Press pp 3-51

 Maydell, B von (1995): ‘Treatment of Third Country Nationals in the Member States of the European Union in terms of Social Law’, in DRICSS (1995) pp 137-154

 OECD (1996): Beyond 2000: the new social policy agenda, Paris: OECD, mimeo

 Pieters, D (1995): ‘Enquiry into the legal foundation of a possible extension of Community provisions on social security to third country nationals legally residing and/or working in the European Union’, in DRICSS (1995) pp 189-244

 Room, G and Perri 6 (1994): ‘Welfare states in Europe and the third sector’, in P 6 and I Vidal (eds): Delivering Welfare: repositioning non-profit and co-operative action in western European welfare states, Barcelona: CIES

 Ruland, F (1994): ‘Impact of international migration on social security: The example of old age insurance in Germany’, in ISSA (1994) pp 77-96

 Soysal, Y (1994): Limits of Citizenship: migrants and postnational membership in Europe, Chicago: University of Chicago Press

 Stephens, J D (1979): The Transition from Capitalism to Socialism, London: Macmillan

 Tamagno, E (1994): ‘Coordination of social security programmes of developed and developing countries’ in ISSA (1994) pp 227-242


ã Copyright 1997. Jean Monnet Chair of European Comparative Politics.

Martin Balwin-Edwards.

European Forum Fellow, EUI, Florence. Co-Director, Mediterranean Migration Observatory. Athens.