Women Migrants and Refugees in the European Union.

The earlier migrants, mainly from Mediterranean countries and European colonies, who entered in the decades after the Second World War, now form settled groups with second and third generations. In the past two decades new patterns of female migration have emerged as a result of geopolitical conflicts and economic restructuring in Eastern Europe and the Third World. So too has international mobility due to economic opportunities, higher levels of education and possibilities for travel amplified temporary an permanent migratory movements. Old colonial links have not lost their force and these flows frequently exhibit high female proportions, for example, Old Commonwealth migration to the UK or Latin Americans to Portugal and Spain, though one of the distinctive features of contemporary migration has been the development of flows to countries with which migrants have not had previous connections, especially Eastern Europeans in Southern Europe. Other groups, such as the Filipinas, have slowly expanded their geographical field in the past decade. The Commission for Filipino Migrant Workers estimated that there were over 500,000 (including families) in the European Union in 1995.

Changing labour market needs and immigration policies are also reshaping female migratory flows. On the one hand, opportunities for skilled female migrants are opening up, particularly in welfare sectors. On the other hand, reluctance to recognise real labour needs as part of a more managed immigration policy, leaves female migrants little scope other than informal work, especially in traditional sectors such as domestic work and the sex trade. As a result, high levels of deskilling and under-utilisation of skills are hidden. Thus diversification, heterogeneity and polarisation typify the beginning of the 21st century (Kofman 2003), although the stock image of migrant women as unqualified and traditional still persists. As Erna Solberg (2002), Minister of Local Government and Regional Development, Norway, tated in her welcome address at the 7th Metropolis conference Immigrants are perceived as homogeneous in public debates, in policy and in research’.

The paper seeks to identify issues specific to migrant women and refugees or which tend to affect them in different ways or on a different scale from male migrants due to gender, racial or other forms of discrimination. Discrimination can be defined “as the use of both different and unequal treatment of a group on the basis of a (real or imaginary) feature or groups of features that are socially construed as negative marks or stigmas compared with dominant players” (ISERES 2000). Migrant women confront legal, direct and indirect discriminations according to their class position, nationality, sexualisation and racialisation. Over time immigration legislation has discriminated less explicitly on gender grounds but it continues to have profound gender outcomes in terms of the preferential categories of immigration and the conditions under which entry, residence and employment are permitted. Yet studies of discrimination and integration have tended to neglect the experiences of female migrants and the second generation (Zegers de Beijl 2000).

The first section of the paper examines the significance of women in the immigrant population and the different ways they have entered European Union countries. Evaluating the size and characteristics of different flows is however made particularly difficult by the unevenness of gender-disaggregated data and the lack of interest in examining gender differences. The second section turns to the participation of migrant women in the labour market, where they are often concentrated in a few sectors. It looks at problems they encounter in selected sectors (domestic, self-employment, community and inter-cultural, health and sex trafficking and trade) and the barriers and obstacles they face in improving and diversifying their position within the labour market. The third section looks at the role of women and gender issues in migrant families and intra-familial relationships. The position of migrant women in the family, especially those from Muslim countries, resonates with broader discussions of their relationship to tradition and patriarchal structures and as vectors of modernity and integration in European societies. The final section briefly raises issues concerning the representation of culture and religion where, on the one hand, new approaches to Tran nationalism and diasporas are situating migrant cultures in more extensive and complex networks, while, on the other, the resurgence of Islamophobia is reinforcing stereotypical images of Muslim women in European societies.

Given the diversification of migratory categories, histories and policies of immigration and integration, there are inevitably major variations between countries in the significance of different flows, participation in the labour market in particular sectors, the size and citizenship of second and subsequent generations, and forms of discrimination confronted b migrant women who are racialised to varying degrees. Though Southern European states have also become societies of long-term settlement of migrants, substantial differences remain compared to Northern states.

Countries without population registers such as France and the U are turning to other sources of annual data, such as national insurance in the UK (Robinson 2002) and residence permits in France (Lebon 2001) to provide more up-to-date and detailed information. Inevitably the numbers of those entering illegally and the undocumented are underestimated. Smuggling and trafficking are serving as institutional networks and diversifying migrant populations in EU countries, for example, the presence of Eastern Europeans in Portugal (Peixoto 2002) and Spain.

Though countries have also developed longitudinal data, such as that based on micro censuses, few are specifically designed to trace migrant trajectories and economic and social discrimination and outcomes. Data based on gender breakdowns at local and national levels are often available but are no thought worth analysing or presenting. A comprehensive report, such as that compiled by Statistics Norway (2002), gives detailed information on migrants (first and second generation and refugees) in relation to demography, education, labour market participation and income. It is one of the few to break down refugees by gender and type of entry (resettlement, asylum). We therefore depend on much of our knowledge of gender divisions and relations in migrant groups on smaller scale surveys or ethnographic studies.

Migrant women have entered the European Union under different immigration categories and for different purposes. Until the stoppage of mass labour migration in the mid-1970s (earlier in the UK), female migrants constituted a significant minority of labour migrants but often entered without children. Following the halt to mass labour shortage, family reunion became the main route of entry into the European Union and was predominantly female. By the 1990s, refugee flows, with variable gender balances began to increase. Quota refugees, such as the Bosnians, often had a more equitable gender balance. From the 1980s Southern European countries clearly shifted from being countries of emigration to countries of immigration, including a strong demand for female labour. Family reunification, initiated by female and male migrants, has also become more significant. The opening up of Eastern Europe and its economic transformation resulted in loss of employment for women and the search for new possibilities in European Union countries. In particular new forms of transient labour migration, often based on a rotational system, enabled women to undertake domestic work, care, cleaning and trading together with familial responsibilities in their home countries (Morokvasic 1996). Others were lured and trafficked into prostitution.

The gender balance differs considerably between sectors. Migrant women have tended to be identified with certain sectors, such as domestic labour, manufacturing, retailing and sex work as well as welfare sectors, such as health. Domestic labour has often been considered as requiring soft skills transferred from the home, and therefore less likely to qualify fo work permits. However, changing evaluations of skills and severe labour shortages are and will alter the gendered balance of labour migration, especially at the skilled end, but also in some instances at the lesser skilled end The implications for female migration will be explored in greater detail in the section on employment.

Family-linked migration remains the main source of permanent migration (estimated at about 65% of permanent immigration for the European Union), and has been particularly dominant in France and Sweden, which have low levels of labour migration. In France, adding together family reunification and the new private and family life visa entries introduced in 1998, two-thirds of first-time long-term residence permits (119,000) were issued for family reasons (Lebon 2002) and thus far higher than the 11% given for work reasons. In Sweden 60.4% of 37, 400 immigrants in 1999 were in the family reunification category (SOPEMI 2001). Yet despite the significance of this form of migration, not just in these two countries (Lahav 1997), and its actual contribution to the labour supply, it receives little attention. In part this is du to its supposedly female dominance (see section on the family) and the assumption that family migrants do not enter the labour force or are not concerned about employment (Kofman et al. 2000). In France, although 80% of family reunification by non-French citizens is female, only about 50% of family formation migration by citizens, many of whom are second generation, is. In Southern European states, family reunification immigration is on the increase. In Italy, for example, 26.4% or 366,122 of residence permits in 2000 were for family reunification (Caritas. 2001). The survival of the household in the country of origin increasingly depends on the livelihood of migrant women (Sassen 2000) who are creating a nexus between the formal and the informal sectors in circuits of counter-globalisation. In Italy, Filipinas, though only the fourth largest group in 2000, sent by far the largest amount of remittances (Caritas di Roma 2002).

Being undocumented and working within the confines of the household makes it difficult to benefit from family reunification procedures. Regularization, as in France and the Southern European states, too may be more difficult for those in informal work and domestic labour to supply proof of employment (Anderson 2000), and virtually impossible for those in the sex trade (Lazaridis 2001). Temporary wor generally does not allow the family to be brought in, but even where it does, as with skilled employment, the conditions of work can preclude bringing in children. Familial separation may leave long-lasting problems for parents and children (Olwig 1999). Caring at a distance imposes burdens on large numbers of migrant women, not just those in full-time employment, especially as ageing parents require more attention or at times when the migrant family needs additional assistance (Ackers 1998). It is exacerbated in many European countries by regulations, which interpret the family in very narrow terms and limit the family members who may join the migrant household (Kofman forthcoming, SOPEMI 2000).

For those entering as family reunification migrants, feminist groups and NGOS have raised a number of issues. These concern the dependency and autonomy of spouses (Kofman et al. 2000), whose residence permits are linked to those of the primary migrant (Freedman and Tarr 2000) and the continuation of their marriage. Although some countries have reduced the probationary period, as in Germany where it was decreased from four to two years in 2001, others have lengthened it from one to two years, for example in the UK, on grounds of the need to deter marriages of convenience. Women marrying men from third World countries are often viewed with suspicion (de Hart 1999). There have been some improvements in the interpretation of the probationary period in that domestic violence, if reported t public authorities, has increasingly been taken into account in deciding the residence status of the spouse. Success in applying for family reunification may be less for migrant women due to their labour market position and greater difficulty in accumulating the necessary resources (income and access to housing). Their presence in the domestic sphere in Southern European countries presents an obstacle for female migrants to bring in male spouses and children. So what offers advantages for women in the beginning may present obstacles once they are more established.

There is no reliable or comprehensive information on the gender balance of asylum seekers and refugees, for although the UNHCR has published statistics on Populations of Concern (women and children) since 1994 (Osaki 1997), it only covered two European states in 2000 (42% in France and 35% in Greece were women but no breakdown by age or marital status was available). Nor do states publish such figures. However Norwegian statistics (Hauge Byberg 2002) by type of entry show that the gender balance has become more equitable since 1990. In terms of principal applicants in 2000, women formed 34% of asylum seekers, 38% of resettlement refugees pre-selected from camps, and 50% of those from the war zones of former Yugoslavia. One of the key subjects of debate is the extent to which women have access as asylum seekers to West European countries and are subsequently able to gain recognition as Geneva Convention refugees or a secondary status such as the Exceptional Leave to Remain in the UK. In relation to the first aspect it is clear that women are less able to reach European countries a principal applicants. The second aspect raises quite complex issues concerning whether their political activities and specific forms of gender persecution, such as sexual violence or behaviour and dress in public, are recognised in the asylum determination process (Wetten et al. 2001). Some argue that women’s political activities, which are often located in the private sphere or involve sustaining dissidents, do not conform to the prototypical male refugee, others contend that women do not fare worse in the determination process (Bhabha 2002). Although Dutch analysis of asylum determination process was not conclusive about gender bias, few of the asylum applications by women in the Netherlands were based on gender persecution grounds. Few European states have developed guidelines for gender persecution in their asylum determination process, along the lines implemented in Australia, Canada and the USA (Crawley 2001). The third area of concern relates to specific problems of settlement and integration, especially the extent to which female refugees have access to and are able to use training, language classes and social resources (see section on employment).

Student migrations are increasingly seen as a form of migration of qualified labour and this has meant that in many countries it has been made easier to switch from student to worker status (SOPEMI 2001). National qualifications are an important factor in success in the labour market (Haque et al. 2002). Shortages in female-dominated professions, such as education and health, may make it more likely that women will be able to remain after finishing their degrees, which tend to be in the Humanities and Social Sciences. Women form 50% of students in OECD countries but the percentage varies enormously according to region of origin. For example, the proportion of women from North African countries in France is low due to restrictions on their movement (Borgogno and Vollenweider-Andresen 1998). Language courses are another form of student entry, expanding the numbers enrolled in educational establishments.

Gendered divisions of labour mean that women and men circulate differently in the global economy. At the skilled end, men overwhelmingly form the mass of those moving within transnational corporations and in the Information Technology and Scientific sectors (HRST), upon which the notion of the highly skilled has been constructed (OECD 2002), and for whom movement was facilitated until the IT bubble burst. Thus 88% of the Green card permits in Germany in 2000 were taken up by men (SOPEMI 2001) The vast majority of scientists from Eastern European in the early 1990s were men, reflecting an uneven gender distribution in Germany rather than an imbalance in the sending countries (Morokvasic 2002).Yet there is little data or information available on gender divisions of skilled and professional labour (Iredale 2001). Skilled employment is, however, far more extensive than the male dominated sectors of management and science (Mahroum 2001). Women, in contrast, have tended to go into what can be broadly classified as the welfare professions (education, health, social work) which have been largely closed (with the partial exception of the UK) to migrant labour (Kofman 2000). An analysis of UK wor permit data for 2000 (Dobson et al. 2001) showed that sectors with high proportions of female staff constitute some of the fastest growing sectors of migrant employment. All professional health occupations and education are currently posted on the Work Permits UK website as priority areas, unlike IT which has been demoted. The ageing workforce throughout Europe in these sectors looks likely to require non-EU recruitment, which for some countries such as Germany will probably come fro Eastern Europe.

Until recently the growth of employment for migrant women was driven by the informalisation of service sectors and inadequacies of welfare provision. Women migrants are prominent in several of the 3D jobs – dirty, degrading and dangerous (domestic, textiles, hotel and catering, agriculture). The search for cheap and docile labour in agriculture has encouraged the use of Eastern European labour, for example Polish and Romanian women who replaced Moroccan men in Southern Spain. A number of those who stayed have ended up in prostitution (Bell 2002).

All these sectors are labour intensive. Unlike in the earlier period of industrial production when women were often employed in manufacturing, migrant women are especially present in the service sector. However, a recent study of migrant women’s economic contribution in Belgium, Portugal, Germany an the Netherlands (ISERES 2000) highlighted that this was often in low qualified jobs, for which they are over-qualified or not related to their qualifications and diplomas. However, having the nationality of the country they lived in did not prevent discrimination.

Most research has concentrated on two sectors : the domestic sphere (cleaning and care) and the globalization of the sex industry where it is estimated that 80% of those trafficked are involved in sex work. All too often female migration is associated with these two sectors with little research being undertaken into their presence in other areas such as industrial cleaning, retailing, hotel and catering, tourism and the self-employed. Women too have used their skills to develop a niche in the community and inter-cultural sector in advocacy, mediation, translation and general community tasks.

In terms of participation rates and unemployment, two broad geographical zones can be identified (table 2). Of course these statistics have to be approached with caution since they under record work in the informal sector and job-seeking registration. In Northern European states, the average rate of participation in the labour force for migrant women (excluding citizens of migrant origin) is lower than for national women but has increased substantially during the 1990s. Refugee women find it particularly difficult to enter the labour market as Norwegian data shows and fewer women participate in labour market schemes than men (Hauge Byberg 2002). A skills audit in London revealed the high level of wastage of their qualifications and the difficulties they have in entering into professional employment (Dumper 2002). In general refugees confront enormous cultural and language barriers, racism, prejudice and lack of recognition of qualifications (Westin and Dingu-Kyrklund 2000), but women refugees have additional burdens of child care and gender stereotyping in terms of suitable jobs, as well as for some, opposition from men to their participation in the labour markets (ECRE 2001, Sargeant et al. 1999). Refugee women with childcare responsibilities find it almost impossible to work.

In Northern Europe, migrant women’s unemployment rates are however substantially higher than national women, in some cases over twice the national level as in Belgium, France, the Netherlands and Sweden. In the four Southern European countries, the participation rate is higher than for national labour due undoubtedly to the large number of independent women migrants. There is however no single pattern for unemployment ranging from a substantially lower rate in Spain, slightly higher in Greece and Italy, and considerably higher in Portugal.

Within each country there are large variations between nationalities. For example in France (Borel et al. 2001), women from Turkey and Algeria have much lower rates of participation (50% and 70% respectively between 20 and 26 years in 1999) and leave the labour market with age ; those from Portugal and South East Asia have a higher or similar rate to French women and also have similar rates of part-time work. Although direct personal services form the single largest sector for female migrants, only 10% of Turkish women work in it compared to 23% of Moroccans and 40% of Portuguese. Vulnerability t unemployment also varies between nationalities in France and illustrates increasing diversity (ISERES 2000). While 24% of foreign women were unemployed, this ranged from only 10% of Portuguese and 16 Italian to 35.8% of African, 39.3% Algerian and 42.5% Turkish.

Although it can be easier for women to find work in personal services for which work permits are allocated in Southern Europe (Italy and Spain), this in turn may lock them into a narrow range of employment. Frequent regularization programmes in Southern Europe have helped many undocumented women gain a legal status and enable them to make more stable and long-term plans, including bringing spouses and children into the country.

Some migrants and refugees use educational facilities to improve their linguistic ability and qualifications. A number of countries have put in place more structured refugee settlement programmes with an emphasis on employment and training, including specific programmes for women. Individuals also devise re-skilling and education strategies whilst they are doing domestic labour, cleaning and hotel and catering jobs (Kleinman 2001). Class and education are major determinants of their ability to regain a lost status and occupation.

Traditionally domestic service was the major employer of women until after the Second World War and this historical process is partially being recreated (Friese 1995), especially in Southern Europe where this sector is the main source of employment for migrant women. Its dominance in Southern Europe an its expansion in Northern European states have rendered the female domestic worker more visible. The participation of indigenous women in the labour force is both dependent upon and creates demand for domestic work which is largely supplied by women of migrant origin, though to a far lesser extent in th UK (Anderson 2000). Domestic work is now part of a global chain of care (Lutz 2002), a globalised domesticity which facilitates participation in the labour force (Ribas-Mateos 2002).

Though patterns of employment, especially the presence of live in domestics, are different between Southern and Northern countries, there are also common issues and problems relating to the nature of the work. The most widespread problems are the low pay and long working hours. Their inferior position and the highly personalised relationship with their employers make it difficult for domestic workers to receive their agreed pay or get time off as agreed. Employers regularly demand unpaid overtime. The health and safety situation in the home is not satisfactory, and if domestic workers are ill, they do not get paid and may even lose their jobs. Psychological, physical and sexual abuse is common.

Domestic work is naturalised as being particularly suited to women who are deemed to innately possess the requisite skills, transferred from one private space to another. At the same time it is work that is not socially valued, and in many Northern European states, denied as valid work for the allocation of work permit. In the UK domestic workers quotas were phased out in 1979 and domestic workers were only admitted as a concession to foreign employers. Workers did not have their own permit or the right to change employers. Over the years numerous cases of abuse and exploitation were reported. Years of campaigning by Kalayaan led to a change in immigration regulations in 1997 and the special right of regularisation for those who remained in the country undocumented (Anderson 2000). The new legislation only allows skilled domestic workers e.g. Nannies or cooks to accompany an employer.

Together with the racialisation of migrant women, the notion of domestic work reifies them in an inferior position, which devalues their skills and portrays them as unskilled and only fit for this kind of work. Yet so many studies of domestic work (Andall 2000 ; Escriva 1997 ; Friese 1995 ; Zontini 2001) highlight the severe degree of deskilling and disqualification that many migrant women with full high school education and even university degrees experience. This applies particularly to Filipinas, Latin Americans and Eastern Europeans. The closure of European labour markets to Third country migrants with professional qualifications (with a few exceptions) meant that for many women the only way of gaining access as a labour migrant was either in the domestic sector or other low level service jobs. The boundaries between the legal and the illegal are often blurred. Many entered as tourists or in some cases as students and became overstayers.

There are of course differences between countries and nationalities. Southern European countries recognise the domestic sector as an area of employment. Spain establishes an annual quota and in Italy the recent legislation (2002) specifically allows for the regularisation of domestic workers who were seen to be less threatening and of strategic importance for the Italian economy ((Fasano and Zucchini 2002). In some Northern countries, for example Germany, it has since 2002 become possible for citizens of accession countries to legally work in the domestic sector for up to three years, though in theory not to undertake caring tasks which could be done by members of the family or qualified carers. Within the confines of the private sphere, this neat division is unlikely t be sustained by employers who are supposed to pa German-level wages. Until now Eastern Europeans have been predominantly pendular migrants (Morokvasic 1996), often rotating job between several people. They have rights of residence for up to 3 months, while others from Latin America and South East Asia are likely to be undocumented. Au pairs from Eastern Europe too are a form of hidden domestic labour which enables them to build up contacts and overstay as undocumented domestic workers. In France, domestic employment is more regulated by the state and still often done by established migrants such as the Portuguese.

Stratification by nationality, religion, race and language skills, leads to different conditions and pay. Filipinas are generally viewed as the most valuable domestic workers, being Christian, English-speaking and well educated. On the other hand, Albanians in Greece or Moroccans in Spain are considered less valuable and have less negotiating power with their employers, often doing less rewarding work an receiving lower wages (Anthias and Lazardis 2000).

The European Parliament (2000) has drawn up proposals to validate domestic labour, to give workers permits not tied to a single employer and to tackle the problem of physical and sexual abuse and exploitation.

For ethnic groups this has been seen as circumventing employment blockages, responding to unemployment and escaping from racialised work places : a way of using cultural resources, networks an skills and serving an ethnic clientele. It has been rapidly increasing in European states. Men far more than women have started up ethnic businesses, although in some countries, such as the UK, minority women are more likely to be entrepreneurs than white women. Women too are jointly participating with their partners in businesses and not simply supplying an unrecognised and unremunerated familial labour force, though this also remains true for many women. Self-employment is far more common among certain groups such as Asians, Chinese and Turkish.

Do women pursue different strategies in entrepreneurship and confront different obstacles ? Several comparative European studies (Morokovasic 1991 ; Anthias 2000 ; Phizacklea and Ram 1996), have shown that women tend to enter into self-employment to achieve flexibility in combining work and familial responsibilities ; they have gained a degree of independence, though often with little support from partners or families (Anthias 2000). Migrant women reported that they were often confronted with indifference or discouragement in their attempt to cross boundaries of what was thought appropriate as female employment, for example the attitudes of bank managers. In the study of five European countries (Morokvasic 1991), self-employment for the majority followed on from salaried employment. The business was not necessarily directed towards services within their own community (Hillmann 1999). A recent Italian study of independent work (businesses, self-employed and independent professional) in the province of Milan showed that just under a fifth of individual businesses were owned by women ; the highest proportions were to be found in fashion production, marketing and socio-cultural services (Zanfrini 2002). The latter is related to increasing involvement in the community and inter-cultural sectors.

Encompassing a range of activities such as advocacy, mediation, interpreting and general community work, this is an area which has enabled migrant women to break out of manufacturing and low level service employment and deploy other skills. Mediation between individuals and families of migrant origin and public authorities and agencies, especially welfare and legal services, have become common in European states (Kofman et al. 2000). Initially, migrant women did it on a voluntary basis and were no salaried. Nowadays, many women migrants and refugees, who are unable to work in the area of their qualification, have found employment in this sector. One of the demands has been the recognition of mediators and inter-cultural specialists as a career (Delacroix 1997) and accreditation of training. This could also facilitate movement into the public and paid voluntary sectors and a stepping-stone into relevant professions. Whilst this could open up more rewarding employment, it might at the same time leave migrants trapped in services for migrants, as has been noted in social work and teaching in Germany and the Netherlands (Lutz 1993). Home language instruction has also presented women with more professional opportunities and has, for example in Sweden, put teaching tenth on the list of jobs done by migrant women (Knocke 1999). Though there will be an expansion of integration programmes in the future, these jobs may well only generate fixed-term and insecure employment.

There has been interest in utilizing the inter-cultural, which could be defined as communication between cultures, as a resource that could be translated into professional competence in a range of fields such as trading, tourism, IT and media, and not exclusively in mediation and advocacy for migrant communities (Federal Institute for Vocational Training. 2000). As with mediation, it should be viewed as a professional qualification and not just a personal feature of female migrants or something they are predisposed to. There is also potential for employment in the creation and management of cultural diversity and tourism for women in professional and entrepreneurial positions, and interfacing with the growth of the knowledge economy, rather than the precarious and seasonal employment usually associated with leisure and tourism. The recent development of this heterogeneous sector means that little research has been conducted into opportunities and gendered outcomes.

Shortages in a range of occupations in this sector, all of which are feminized to varying degrees, are beginning to occur in European states. These shortages at different levels of skills and qualifications are providing more opportunities for women to migrate officially for short and long periods from developed and developing countries. In particular the crisis in nursing and recourse to overseas nurses within an increasingly globalized labour market has received much recent attention (Buchan 2002 ; Kingma 2001 Tjadens 2002). International recruitment occurs throughout Europe but with the exception of the UK, and more recently Ireland (Yeates 2003), with a more global reach, has primarily looked to neighbouring countries, as in the Netherlands. Opportunities for nurses are likely to extend to other European countries, for example in Italy which in 2001 included a quota for 2000 professional nurses in its migrant labour force programming (Zanfrini 2002). The global treasure hunt to meet shortages has generated discussions about guidelines for ethical recruitment and the redistribution of the crisis in personnel to developing countries, which have not been able to retain their skilled workforce.

Since 1995/6 the number of overseas nurses in the UK has risen dramatically and the country is diversifying its range of source countries away from its traditional Commonwealth suppliers. Active recruitment, for example from the Philippines, is facilitating a rapid increase of nurses from the Third World at a time when intra-European registrants have stagnated. Data from the UKCC/Nursing and Midwifery Council on registrants and work permits (new and extended) reveal that the biggest increases have occurred in the past few years from the Philippines (7422 work permits granted in 2001), India (1759) and South Africa (1163). The conditions that lead nurses to leave their home countries (low remuneration, heavy workloads, limited training facilities, poor career development, resource shortages and personal safety) are likely to persist. So too are some of the reasons for shortages in European states : an ageing nursing force caring for an elderly population, inadequate numbers being trained in the early to mid 1990s, alternatives available for a traditional female work force and an under-valuation of its worth, partly reflected in low pay. It is therefore likely that nursing will continue to provide employment opportunities for migrant women from the Third World, Eastern Europe and other developed countries. The continuing demand for nurses and carers may have an impact on patterns of recruitment of domestic labour, especially in relation to Filipinas who figure prominently in both sectors.

Compared to the older forms of incorporation into the non-career grades of skilled labour markets, current shortages mean that nurses are not being inserted to the same extent into the lower and subordinate ranks as earlier migrants from the Caribbean and Ireland were. However, in the UK the conditions of wor vary between the public sector and private nursing homes, where stories of exploitation have been most common.

Though the most publicised at a European level, nurses are not the only area of severe shortage. In al the health occupations in the UK in 2000, the foreign-born, the vast majority being non-EEA-born, occupied a significant place (Dobson et al. 2001). Overseas doctors (non EEA- qualified) for example form a large percentage of the UK medical labour market ; in 2000 they constituted 26% of the hospital medical workforce. Most women had come to obtain further qualifications and training, and were, as with male doctors, in the middle and lower grades. In an increasingly feminised occupation, women form significant proportion migrant doctors with over half of those seeking registration with the General Medical Council in 1998 being female (Raghuram and Kofman 2002).

Care involving children, disabled and the elderly is also a growth area and often takes place in the home or in private nursing homes. This development may represent a step upwards for those in domestic labour, enabling them to use nursing skills and earn better pay. Qualified care and other health assistants provide employment for both recent and more established minorities. In some countries as in Germany, integration measures are seeking to utilise employment in this area as a means of improving the skills and labour market positions of established minorities.

Reclassifications of tasks and skills may also alter the exact nature of shortages and the use of migrant labour. Nurse practitioners are beginning to take over some of the work of doctors eg. writing out prescriptions, while health care assistants are taking on some of the more basic tasks done by nurses.

Its increase in the 1990s, especially from Eastern and Central Europe, which has to a great extent replaced flows from Latin America and Asia, is related to the globalisation of the sex trade. The commodification and trading of bodies, aggravated by the feminisation of poverty and inequalities, have created a form of female slavery (European Commission 2001). Trafficking has become increasingly territorial and sex markets controlled by organised crime. It is estimated that between 200,000 and 500,000 are working illegally as sex workers within the European Union. Trafficking has been defined by the UN Protocols as the recruitment, transportation, transfer, harbouring or receipt of persons by means of threat or use of force or other forms of coercion ; it is differentiated from smuggling which refers to gain derived from the illegal transport of persons across state borders. In the case of women it may be difficult to distinguish between the two. Discourse on trafficking by recipient states has been couched in terms of migration control and organised crime, while NGOS have tended to focus on human rights and the abuse and welfare of trafficked women (Ucarer 2001).

Although there is increasing literature and concern about the subject of trafficking, there are still few systematic and reliable statistics available for comparison. Germany, however does collect data over time on the victims of trafficking discovered in the country but these only cover trafficked women involved in police investigations. In addition to official data, a number of NGOs and IOM have produced in-depth studies and statistics. On the basis of statistics (produced by NGOs) in Western European countries in 199 and 2000, trafficked women largely originated from the Baltics and CIS (19.7%), Central Europe (5.8%), Balkans (20.8%), Africa (28%) and Other/Unknown (25.1%). Eastern European countries, especially Belarus, Lithuania, Russia, Ukraine and Moldova, have become sending, transit and destination countries. There is evidence of chain trafficking and privileged links between countries, such as Lithuanians in Germany or Nigerians in Italy. In particular, the lack of economic opportunities and poverty play a major role.

Three types of networks have been identified : the large scale based on structures of international contacts using a variety of recruiting methods and often passing through transit countries ; medium scal which does not sell women to other groups but keeps the women under its control ; and the small scale operating with recruitment at an individual level and in contact with say a club owner in another country. In this case the sex worker may well have a close rapport with the recruiter. In Greece only about 20% knew they were going into sex work, many drifted into it or used it to supplement domestic work (Lazaridis 2001).

Most of the women trafficked had, as single mothers and divorcees, sought jobs in Western Europe to support their children (Ucarer 1998) or help parents without financial support. A general improvement in the economy and more opportunities for women, who have been hard hit by the marketisation of economies, as well as greater awareness of the risks of trafficking, seem to have contributed to a decline of trafficking in Central Europe, as in the case of Poland.

Most of the EU and international initiatives since 1996 have aimed to control trafficking and more recently to give some protection to victims who cooperate but this fails to take into account the demand for sex workers in the destination countries. The sex trade has become a lucrative business (Salt 2001), fuelled by the growing demand in destination countries for foreign and exotic prostitutes (IOM 1996). Some (Taylor and Jamieson 1999) argue that we require a more complex analysis than that based on organised crime and have to take into account the liberalisation and mainstreaming of sexual representation within a market culture. They also suggest we need to understand the context of recruitment of women through the media and the internet as well as the consumption of sex. The upshot is that the demand for sex workers is embedded within contemporary cultural production, and policy dealing with sexual trafficking cannot be solely directed towards organised criminal groups.

One o the major disagreement between the EC and NGOs is over whether prostitution should be legal and whether efforts to prosecute it are a violation of human rights (Ucarer 2001). Prostitution itself is regulated and covered by law in very different ways in European states (Femmigration. 2002). As Campani (2000) and Lazaridis (2001) show for migrant prostitutes in Italy and Greece, the conditions of work, degree of possible autonomy and exploitation vary according to the type of prostitution. In Italy those whose prostitution is masked by other activities such as dancer, strip tease, masseuse, hostess etc, and take place in private and spaces, are more likely to be from Eastern Europe and Latin America, they generally have more autonomy and are older than those practising in the street. The latter is largely undertaken by women who have been trafficked and controlled by organised criminal networks. This applies especially to Albanian and Nigerian women. Campani concludes that quitting prostitution often depends on friends or family and NGOs (Council of Europe 1998) providing assistance and support. At the same time, one has to take into account the fact that not all women have been forced into sex work by traffickers. There is therefore a need for multi-disciplinary approaches in research and policy initiatives on this topic (Ucarer 2001 ; van Impe 2000).

The concept of « migrant families » is still neglected in European scholarship on migration processes and strategies (Fernandez de la Hoz 2002), unlike research being conducted on transnational migrations within the Asia-Pacific region and between North and Central America (Kofman forthcoming). One of the contributing factors to the neglect of the family is that in economic theory transactions are considered to occur between the individual and the state (Vatz Laaroussi 2001, Zlotnik 1995). Lopes et al. (1994), writing about the failure to recognise the significance of the family in European Community policy, argued very strongly that family mobility represents the interface between the individual and the social and of public and private spaces

Family migration is treated in policy terms as a secondary type of migration, which is viewed as an unintended consequence of the stoppage of mass labour migration in the 1970s, and consisting of female dependants following the male breadwinner as the primary migrant. A further unwarranted assumption is that women do not seek to participate in the labour force or, if they do, only in order to earn additional pin money. This tends to result in a disjuncture between research on migrant and minority women in the labour market and their role in the private sphere, often as upholders of traditional cultures. Where policies for the adaptation and integration of migrants were implemented, as in France, such perspectives meant that programmes were developed for basic language skills and how to be a housewife in a modern society rather than access to the labour market and training. The lack of research on family reunification and marriage migration amongst second and subsequent generations (Kofman forthcoming) has yielded only disparate evidence about the increasing educational background of more recent migrant women entering through family reunification (Borel et al. 2001, Tribablat 1995) and their individual and collective projects.

Earlier studies of European migrant families focused primarily on those whose cultural distance from mainstream European families was the greatest ; migrant families were represented as being caught between two cultures (Vatz Laaroussi 2001). Asians in the UK and North Africans and Turks in Belgium, France, Germany and Netherlands captured the attention of researchers and policy makers (Boubeker 1999 ; European Observatory on the Social Situation, Demography and Family 2002 ; Nauck 1994). Most, though not all of these groups were Islamic, and therefore lent themselves to representations of women imprisoned in the private patriarchal domain. The difficult voyage to modernity for women was seen to begin with their entry into the country of immigration as if the society of origin and gender relations were fixed in time (Kofman 1997). By the 1980s, some feminist researchers entered into the private world and spotlighted women as agents of changing gender relations as families evolved in the process of migration (Boulahbel-Villac 1994). One could observe women taking major decisions about their children’s education in Algerian families in France. Through their children women had more contact with the external world of social and welfare services and were participating in more religious tasks in the making o symbolic life (Andezian 1986).

One of the issues in recent years has been the marriage strategies of young women of North African and Turkish background in France, Belgium and the Netherlands (Boulahbel-Villac 1994 ; Lesthaeghe and Surkyn 1995 ; Lievens 1999). Much of this literature rejects the notion of a simple shift from tradition t modernity and highlights the process of negotiation between the two poles. The modern nuclear family, often attained through migration, is a major step from the extended family and social control imposed in the country of origin. Marriage may be a means of gaining a degree of freedom, and women are more likely to migrate for social emancipation, especially from strongly patriarchal societies. In groups where arranged marriages were previously the norm, its meaning and the degree to which it continues to b practised have also changed, especially for younger and more educated women. A study of family formation in the UK highlights changing attitudes towards choice of partners for younger women in different South Asian groups (Berthoud 2000), particularly for those who were born in the country or had migrated while very young. It also shows that the rejection of arranged marriages has progressed most rapidly amongst Hindus but to a lesser degree amongst Muslims and Sikhs. East African Asians, on the other hand, a group that had previously migrated and largely came from an urban trading background, had patterns that were very similar to the indigenous population in terms of the percentage of women who participated in the labour market and whose husband had the final decision at home. These findings demonstrate the significance of age and previous socio-economic background (Bhachu 1991).

Despite the emphasis on migrant families of established minority groups, or those with ‘problematic’ cultural and social customs such as West Africans in France, who transported polygamous familial structures to France in the 1980s, the diversity of forms and practices of migrant families in European states are being recognised. In the 1990s, rotational migration from Eastern European countries ((Kupiszewski 2001, Morokvasic 1996, Okolski 2001), and in particular female labour migrants, many o whom were initially undocumented, left behind family members (Phizacklea and Anderson 2002, Zontini 2002), thus reproducing a pattern of separated families that had been common amongst many nationalities in the earlier guest-worker period (Charbit and Bertrand 1985). Caring at a distance involves children an other relatives ; the “form such practices take is determined by the socio-economic conditions and normative aspects which underlie migration” (Baldassar and Baldock 2000). Migrant women, who have entered as labour migrants (Philippines, Latin America, Eastern Europe) and have children, are forced to juggle work and care. This may lead them to seek jobs with flexible hours or non-live-in domestic work. Others may leave their children behind with kin in the home country, and this applies not just to those who are undocumented, but also to skilled women on temporary contracts, such as Filipina nurses in the UK. Immigration policies make it even more difficult for many migrant families to have recourse to other family members, such as parents and siblings when they most need them.

Hence the relationship between family structures, gender relations and presence in the labour market of migrant women needs to take into account a number of factors, such as mode of entry into the country, family relationships and decision-making, familial responsibilities and access to care services, educational levels, age, and migratory trajectory. Although much research is being carried out on indigenous families and gender relations, current reviews of migrant families (Fernandez de la Hoz 2002, Pflegerl 2002).demonstrate how under-represented gender issues are and how much research is needed to acquire a deeper understanding of the different strategies deployed in migrant families by women, men and children. Such research should not assume the nuclear or extended family as the norm amongst migrant and minority groups but include single parent (a substantial percentage of Afro-Caribbean households are female-headed) and same sex families.

There is some excellent and nuanced research in case studies on employment, decision-making, gender and generational relationships that complements evidence from large-scale surveys. Studies of Pakistani and Bangladeshi families in the UK (Dale et al. 2002, Evans Lloyd and Bowlby 2000) note that the decision to enter employment in the public sphere depends on educational level, family resources an class position. For married women, the decision is often negotiated with parents-in-law as well as the husband. On the other hand, French research with Algerian women (Boulahbel-Villac 1994) showed that though working at home, some women had devised ways of earning money of their own by making clothes and jewellery.

In the UK and France, there has been considerable research on migrant and ethnic minority youth (Migrations Société 1999a), usually focussing on the major groups such as Afro-Caribbean and Asian in the UK and North Africans in France. In Southern European countries, however, there are still virtually no studies of second generation young people (but see Andall 2002). An EU project on the integration of migrant children in seven European countries (EFFNATIS) has raised the issue of the role of parents and their own educational standards in the educational achievement of children. Large-scale surveys have shown the relationship between class background, cultural values and educational attainment but have not necessarily probed the aspirations of parents and children and how these vary by gender. Studies have often shown that girls perform better than boys because of additional restrictions placed upon them and their desire to gain qualifications as a means of acquiring some independence. Dale et al. (2002) found that parents in Bangladeshi and Pakistani families (the most disadvantaged minorities in the UK) supported the idea of girls’ education, although it could be frustrated in families with traditional expectations. We should be cautious of equating traditional attitudes to the education of women with Islamic families treated as a single entity. A study (Crul 2000) of educational attainments of Moroccan and Turkish second generation in the Netherlands found that Moroccan girls performed well whilst Turkish girls were under extreme pressure to leave school. It was the strength of the ethnic community, social relations with Dutch people and the time of entry into the country which explained a good part of the difference between Moroccan and Turkish girls. In a study of female students from North Africa in France (Borgogno and Vollenwider-Andresen 1998), the educational background of their parents was found to be significant in whether they were allowed to travel abroad to acquire qualifications.

Given the diversity of migrant groups and their gender and power relations in Europe, discussing culture, religion and migrant women is fraught with difficulties One can only mention some general trends. In the 1980s many of the discussions about the relationship of migrant women and cultural practices were played out through the dichotomy of tradition and modernity and their potential as vectors of integration, as they were in their role in the family. Reproduction of culture through the family was central to the maintenance of the collectivity (Anthias and Yuval Davis 1989 ; Yuval Davis 1997), frequently putting pressure on women migrants to conform to traditional practices.

The notion of two monolithic culture blocs, between which young women navigate, has increasingly been rejected. It is more appropriate to speak of two cultural heritages shared to some extent intergenerationally b mothers and daughters (Lutz 1995). The significance of gender in the construction of hybrid identities stemming from transnational and diasporic networks and in the context of multicultural societies in Europe (Modood and Werbner 1997) is being addressed (Brah 1996 ; Purwar and Raghuram 2003 ; Yuval Davis 1997). As Bhachu (1991) argued for Punjabi Sikh women, specificities of class and ethnic cultural values govern the production of particular identities in specific contexts as with white women. Migrant women too are representing culturally their own gender identities and communities through cinema, literature and art (Freedman and Tarr 2000).

However the strengthening of Islamophobia (Runnymede Trust 1997) and emphasis on cultural practices (head scarf, arranged marriages, honour killings) associated with Islam focuses attention more than ever on Muslim women oppressed by patriarchal systems (Blaschke, J. et al. 2001 ; Dietz and El-Shohoumi 2002). Even before the Gulf War, the link of Islam with terrorism and the oppression an expulsion of women from the public sphere in Afghanistan, the headscarf affair in France in 1989 had propelled Muslim girls into the limelight (Dayan-Herzbrun 2000). Apart from the veiling of Muslim women in public places, two extreme practices, that o forced marriages and honour killings, have captured much media attention which has often portrayed Muslim women as unrelentingly oppressed by dogma and without any religious autonomy (Patel 1999). Such practices have also generated official debate and policy measures, as in the UK, Denmark, Norway and Sweden. Forced marriage is defined as a marriage contracted without the free and valid consent of one or both parties (Southall Black Sisters 2001). It is different from arranged marriages where there is free and valid consent, although in reality the difference for many women may not be so clear. As more recent studies of educational achievement and familial relations have shown, shifts in gender relations in the workplace and the household intersect with and are shaped by class and education. Downward mobility, racism and discrimination contribute to conservative familial and gender identities. An understanding of how young Muslim women, like other women, construct their identity through values and norms that may be both contradictory and complementary (Moulin and Lacombe 1999), needs to take into account the double discrimination of sexism and racism in relation to the wider society and the internal communal pressures they confront (Hashimi 1999). Muslim women need to be given their own voices (Diamantopoulou 2002). Article 13 of the Amsterdam Treaty will cover discrimination on religious grounds but that will have to counter the heightened racism against Muslims since September 11t 2001, especially against those who visibly affirm their religious affiliation, as with veiled women in the workplace and in schools.

In Europe, women and migration were for a long time associated with family reunification. Progressively the reality of migrant women in and out of the labour market in Northern European states questioned this relationship. Their presence in Southern European labour forces, even more than in Northern European states, has drawn attention to women as independent labour migrants and initiators of family migration. Today with the increasing severity of shortages across numerous sectors and an ageing population, the presence and demand for women in skilled sectors must also be taken into account. At the same time the continuing informalisation of economic processes in European states, and the precariousness and feminisation of poverty in Eastern Europe an Third World countries, feed a growing undocumented population in a range of services.

Diversity and polarisation are increasing. The polarisation of employment conditions, rights an citizenship is likely to be marked in the UK and Germany which have opened up a number of their skilled sectors to migrant labour The skilled will be able to convert their mobility into settlement and eventual citizenship ; the less skilled, though not necessarily the less needed, are being offered at best temporary contracts with no right to bring in family members. Though applicable to women and men, the playing ou of this accrued polarisation has distinct gendered aspects. The feminised global chain of care increasingly traverses both the skilled and valued and the lesser skilled and devalued. In this context it is worth asking how new European anti-discrimination legislation will affect migrant women positioned differently in relation to immigration legislation and its creation of new employment opportunities and stratifications.

What is also clear is that we cannot capture the nature of female migration by analyses that maintain the disjuncture between the economic and the social, between labour processes and family life. The sections on employment and family life have highlighted their interaction. Women may migrate to ensure the survival of their families but at the same time pursue individual projects. Migration may thus bring self-fulfilment as well as enslavement. And as we have seen, those who are skilled within increasingly globalised labour markets may not have to face this tension. It is thus imperative that policy measures recognise the growing polarisation and differentiation and moves beyond an exclusive focus on the negative aspects of gendered migrations.

The paper has identified a number of areas which require further statistical analysis and research to understand the gendered nature of immigration and integration and policy initiatives to improve the conditions of and opportunities for migrant women and their children.