• 1  Only 43% of those surveyed in France agreed that there were “too many foreign people in France,” w (...)

1In the past five years, immigration policies of major European Union governments have come under fire from within. Heads of State in France, Germany and the UK have all taken an increasingly tough tone in their public pronouncements, using “immigration reform” – meaning, anti-immigrant policies and border-closing strategies – as a key part of election campaigns. Their harsh rhetoric has apparently found an echo in public opinion in each country: in a Financial Times/Harris Poll from February 2007, a majority of those surveyed in Great Britain, Italy, Spain and Germany said that their countries contained “too many people from foreign countries”1 (Harris, 2007). In that same poll, a majority in each European country surveyed agreed with the statement that their government should “tighten up its border controls to stem the flow of workers immigrating from Central and Eastern Europe.” The poll asked broad, philosophical questions about the acceptability and propriety of current immigration trends, but shied away from asking questions that were specific to the lives, and quality of life, of middle-class Europeans. For example, respondents were not asked, “Do you approve when someone hires an au pair from Poland?” Nor did they expect respondents to answer the question, “Would you hire a house cleaner from Morocco?” The survey highlights the xenophobia immigrants often face, but also points out the disconnect between government policy and economic and social reality. Are people prepared to clean their own houses, care for their own children and elderly parents, shoulder the burden of reproductive labor on a household-by-household and day-to-day basis? And what would happen to productivity and labor-force participation if they did?

• 2  Increase in undocumented work over time may also weaken the labor movement and erode those protect (...)

2Calls for tighter borders make little sense, and will have little effect, when restrictive laws are not enforced and administrative policies are not adjusted to reflect the new realities of employers’ demand for labor. This tacit acceptance of the inflow of undocumented workers perpetuates the invisibility of immigrants with regard to labor policy; it denies immigrants important safety and security protections granted other workers.2 But it also perpetuates (and is perpetuated by) the invisibility of women in the receiving countries as professionals and full participants in the labor force, and sustains (and is sustained by) the social silence about what Arlie Hochschild has called “the stalled revolution in the family” (Hochschild, 1990).

• 3  By “shadow economy,” we mean economic activity outside the formal sector of employment, activity t (...)

3Our paper examines the current immigration “policy gap”: the combination of restrictive policies with the fact that those policies do not work, because they are unenforced and perhaps unenforceable. Our first level of finding (which is hardly new) is that restrictive policies are at best irrelevant to the flow of workers across borders to take low-paying domestic jobs, often as part of a “shadow economy.”3 We suggest further that restrictive policies may have the opposite effect from the official intentions of receiving governments. What the policy gap does is to keep wages low and workers vulnerable, thus ensuring that they will be attractive to employers. Given the persistence in global income inequality, women in developing countries will still migrate to do domestic work. The policy gap is thus self-perpetuating.

• 4  Bridget Anderson (2007) finds that employers of domestic workers are looking for “foreign” women w (...)

4Feminist analysis is crucial because it can help scholars and policy-makers better understand the demand or “pull factors” within the receiving countries. Domestic work is different from other sorts of labor because it occurs within the so-called “private sphere” of the family, which is both an economic and an ideological formation; and thus it participates in the complexities, contradictions, and mystifications about gender and power which feminist theory has long identified there. But feminist methodology will be incomplete and flawed unless it takes the sort of “intersectionalities” approach recommended by (among others) Chandra Mohanty and Kimberley Crenshawe, which would see gender relations as crucially structured by differentials of race, nation, and ethnicity. Domestic work (whether paid or unpaid) has long been socially defined as “women’s work”; that paid domestic work is increasingly also defined by national and racial distance and difference between employer and employee is no accident. In other words, the demand which drives (and is driven by) the policy gap is not simply for workers who are willing to accept low wages and poor conditions, who just happen to be immigrants because of economic differences between receiving and sending countries (though that does help explain why such workers are available and willing to move.) Rather, the demand is specifically, and crucially, for foreign women to do these sorts of work, and it arises within a web of economic and ideological factors.4

• 5  Dal Lago, 1996 argues that the influence of racism should no longer be considered “American except (...)
• 6  Chandra Mohanty, “Cartographies of Struggle,” in Feminism Without Borders (2003). The intersection (...)

5Domestic work is in-demand everywhere. So domestic workers ought to be very mobile, like nurses or computer programmers. And they are indeed very mobile. But unlike nurses or computer programmers, they are not in a good position to negotiate high wages and good working conditions. Why? Xenophobia, racism, the perception of immigrants as extracommunitarian “Others” certainly play a part, but they are not the whole explanation: there is something peculiar about domestic work that requires a gender analysis.5 At the same time, it is clearer than ever that appealing to some vague notion of women as held responsible for “reproductive labor” and as therefore having some form of oppression in common really does not correspond to demographics or lived reality. An intersectionalities perspective is thus crucial. The gender analysis doesn’t “work” without the race-and-nation based analysis, or vice versa, because the two forms of disadvantage are not additive, they are inextricable: they “mean” together, so they must be “read” simultaneously.6

• 7  Schwenken (2005) uses gap analysis to argue that there are more opportunities for political action (...)

6Scholars studying immigration policy have often noted the discrepancy between the apparent intent of official state policies and the outcomes of these laws and regulation regarding the volume and type of immigration.7 This disparity between aims and effects of immigration policy has been variously attributed to poor planning and complex circumstances that produced unintended results (Martin, 2004; Hollifield, 2004), intermixed with “spotty enforcement” (Calavita, 2004:376). But the underlying reasons for this combination of unattainable policy goals and slack enforcement in a wide range

• 8  See Massey (1990) and Cornelius and Rosenblum (2005) for good overviews of the social and economic (...)
• 9  In particular, see the UN Report: “Specific Groups and Individuals: Migrant Workers.”
  http://www.u (...)

7of countries lie in the contradictions arising out of interest group politics, domestic and international institutional structures, and the underlying social and economic forces that generate migration pressures.8 Amid the complexity and lack of policymaking transparency, immigration policy toward domestic workers, typically women working in other people’s homes, can perhaps be viewed as having the biggest policy gap: because the work is “hidden” in the informal sector and in the private sphere, it has never been properly addressed by policymakers. Until recently, European governments hadn’t addressed migrant domestic labor at all as part of an overall migration management scheme.9

• 10  This becomes even sharper in cases where multiple ethnicities, or perhaps it would be better to sa (...)

8Female migrants have little access to formal sector jobs, and therefore take on informal sector jobs – what Anderson calls “the three C’s”: cooking, cleaning and caring – typically those derived from traditionally delineated gender roles (Anderson, 2000; Ehrenreich and Hochschild, 2002). Migrants will be more able and willing to “live-in” since they have fewer in-country private relationships or personal connections outside the home (Akalin, 2007). Iffy immigration status makes migrants even more vulnerable and dependent on their employers (Anderson, 2007). Moreover, “foreignness” or being ethnically “other” gives employers reasons for hiring migrants over existing local workers: employers may feel morally superior for helping out someone from a poor country or may feel more comfortable having someone not of their own race in a market transaction carried out in their own home (Anderson, 2007). Ideas that in-migrant groups, or different races, have an incompatible gender structure, ideology, or a “different” family, may also serve as an excuse for unconcern about who is looking after the childminder’s children or cleaning the housekeeper’s house.10 In particular, this may address, or at least mask, socially induced anxieties among professional women about whether they are also adequate mothers (and thus adequate women): if the tasks of caring can be done just as well by someone else, what remains of the maternal role and identity? But a paid caregiver whose difference from the child is ethnically marked is less likely to be mistaken for the child’s “real” mother by outsiders (or indeed by the child). The worker needs to be somehow marked as “Other” so that “women’s work” can be distinguished from “woman’s role,” and so that educated women may free themselves from the burden of the former without paying the social costs of being seen to abandon the latter.

• 11  A third reason for studying domestic migrant labor is that it may illuminate long-standing, and de (...)
• 12  Freeman and Schettkat (2005), using standard economic reasoning, argue that workers in the EU shou (...)

9Migratory domestic labor has been of particular interest to feminists for a variety of reasons. First because (like sex trafficking) it puts certain women as women in positions of social disadvantage and serious risk; but second because, given that domestic labor as such remains ineluctably gendered, it appears to put different groups of women at odds with each other, with the upward migration (so to speak) of women in host countries dependent on the geographical migration of women incomers.11 A rise in income levels, and public status, for some women appears to depend on keeping other women poorly paid and invisible. On this account, professional women need caretakers for their children (and sometimes their parents), along with housecleaners to carry out domestic chores in order to compete with men in the workplace12; hiring unauthorized migrant workers as domestics is cheap, with penalties for doing so rarely imposed on employers; in addition, an aging population increases the demand for home health care workers (and the erosion of the welfare state and social responsibility for care makes this situation worse). For a wide range of countries, this problem continues to be seen as a “women’s issue,” because domestic work is “women’s work”; thus the situation is seen as women’s fault (and as women’s responsibility to solve within the so-called private sphere).

• 13  Hochschild, A. R. (2000) “Care Chains and Emotional Surplus Value,” in Hutton, W. and Giddens, A. (...)

11As Hondagneu-Sotelo has observed, the “New World Domestic Order” entails paid domestic work becoming, surprisingly, a “growth industry,” with migration fueled by economic inequality both within the U.S. and European countries and between different countries, leading in turn to what Hochschild has called “global care chains.”13

• 14  See Pissarides, et al. (2004:74) This result is true for all the European countries with the excep (...)

12Women entering the professional ranks in industrialized countries have fueled demand for domestic workers, thereby pulling in women from around the world to clean, care and cook. Table 1 below shows the growing proportion of women in the labor force in various European countries that coincides with women being more likely to work and earning more per hour on average. Education levels were rising as well during this time, with over half of women in the 25-34 age group having tertiary/university education by the year 2000.14 In the highest education category, European women tend to have very high employment rates, ranging from 74% in Spain to 93% in Portugal (Pissarides et al., 2004:73). Many of these women were likely in high income, dual-earner households, the type that would be most likely to hire domestic workers and/or nannies. Not only do these trends explain the persistence of income inequality between European households, but they also point toward the reasons that paid domestic work has been a growth industry: the combination of increased education and higher employment rates as women become a larger percentage of the overall labor force indicates the source of strong demand for domestic service workers’ labor.

• 15  For example, see Barile et al. (1994), Anderson (2000), Parreñas (2001), Akalin (2007).

13From the “supply” side of the equation, we know that there are various family pressures and social forces that may encourage women to migrate: a lack of good job opportunities at home is only part of the issue.15 In recent academic literature, much has been made of migrants’ use of social capital to help them make migration decisions (Massey and Aysa, 2005). Leblang, et al. (2007) find that social networks lessen the risks of migration, with large loose networks helping to provide migrants with useful information and smaller, tighter networks providing better support. Ethnic social networks can also assist immigrants in integrating into the host culture or generate ethnic enclaves, where integration into the larger society is not necessary in order to function as a domestic worker. While these networks may be more or less active in various places, it is clear that these networks commonly enable domestic workers to migrate. As such, the operation of these networks affects how and whether government policy has the intended effect. For example, if the government decides to restrict the number of asylum seekers and limits their ability to work, migrants are more likely to choose to enter illegally with the help of an ethnic network and work in someone’s home, rather than get government assistance as a refugee. The shadow economy serves as an escape valve when other avenues of migration are cut off, and the policy gap widens as a result of the restrictive policy.

• 16  Papademetriou “guesstimates” that there are perhaps 7-8 million irregular migrants in Europe. (See (...)

14Given the private nature of the work, hard data (even reasonably good estimates) on the total number of domestic workers are difficult to come by, but it is safe to say that at least several million migrant women are working in domestic service around the world.16 As a country with one of the largest emigration flows of women, Parreñas (2001:1) argues that from the Philippines alone there may be more than two million female migrant domestic workers spread around the globe. Overall worldwide flows are large, but what is necessary here is to show what is happening in Europe – that this phenomenon is important in a range of European countries that have differing immigration policy. The use of the data here has to be more subtle because of the lack of availability of definitive migration data disaggregated by gender, as noted by Morrison, et al. (2008:2). But this notion that domestic work as a growth industry can be seen in the evidence showing that a larger share of the world’s migrants are going to developed countries, including those in the European Union, and that an increasing proportion of them are women. Since a greater proportion of female migrants are going to Europe, it is likely the case that a greater proportion – both increasing over time and a greater share than to the rest of the world – of domestic workers are going there as well.

• 17  See Eurasylum’s website for full interview, accessed 15 October 2007, available: http://www.eurasy (...)
• 18  Ibid. Also see Chou (2006) who delineates changes necessary in the European Union’s institutional (...)

17Based on various estimates of irregular migration and on the experience of countries that have carried out periodic “regularizations,” we know that the numbers of undocumented workers in European countries are high, with Germany and the UK each having more than 1 million irregular migrants. (See Table 3 below.) It should be noted that these are based on very rough estimates for irregular migrants, and are conservative relative to Papademetriou’s (2005) “guesstimate” of 7-8 million noted earlier. In fact, in a recent interview with Eurasylum, Hervé Carré, Director-General of Eurostat, the statistical arm of the European Commission, stated that the European Union would not even try to estimate the number of irregular migrants in European countries.17 The main goal of the European Union with regard to migration appears to be gathering data and harmonizing statistical procedures, indicating just how distant a “unified” immigration policy may be.18

• 19  These data are based on Papademetriou et al (2004), Table 1. In 2002, Italy’s program aimed at dom (...)
• 20  Ibid, Table 1. It should be noted that regularization sometimes involves receiving work permits fo (...)
• 21  Ibid, Table 1. Germany and the UK allow more workers to come in as refugees or asylum seekers, but (...)

19Looking at the history of large-scale “regularizations” carried out by the major European migrant destinations reveals a pattern of relative leniency in some countries, and greater restrictiveness in others. Since 1980, Italy has regularized roughly 1.5 million workers in five waves of programs, including one in 2002 designed to target caretakers and dependent workers.19 Over the same time period, Spain has regularized over ½ million workers, while the UK regularized fewer than 1000 domestic workers.20 Germany regularized about 30,000 workers during the 1990s21. More restrictive policy does not appear to coincide with reduced flows of migrants, or of the estimates of migrants with irregular status.

• 22  Ibid, p. 15.

20Based on the evidence above, it appears that there are no striking differences in the immigration policy gaps or outcomes based on different government actions or socio-economic contexts. Different immigration policy strategies among large European countries over the past two decades led to similar results: a policy gap that does not depend much on the various contexts or particularities of the economies involved. This finding leads us to agree with Parreñas in her study of Filipina domestic workers in Los Angeles and Rome (Parrenas et al., 2007). She expected to find striking differences in these women’s experiences based on “contexts of reception.” Instead, she argues that the reason for their striking similarities “rests largely on their positioning in globalization as part of the secondary tier labor force of the economic bloc of postindustrial nations.”22

• 23  On this point, see Lillian Trager, ed. Migration and Economy: Social and Local Dynamics. Society f (...)

21But what is wrong with migratory domestic labor? In fact, this does not go without saying. If migratory domestic labor is damaging, we need to lay out clearly how, and why, and for whom, before we can ask at what level solutions need to be proposed. When we refer to a “policy gap,” for instance, our concern is not to help governments design more successful restrictive policies in the interests of racial and national “purity.” On the face of it, connecting workers who need more money than they can make without leaving home with work that needs to be done sounds like a good idea. Moreover, the long-range view taken by historians and anthropologists reminds us that migration is normal, has always been part of the cycles of human life.23 Should we really regard economic migration as a sign of crisis (in either sending or receiving countries), or can we see it as part of stable systems of exchange, as fulfilling a function for both parties (because otherwise it would not be occurring)? It is possible that those countries that are attempting to restrict immigration are acting from simple racism or even simple fear of change, and they should be told to “get over it.” Especially since in the case we are discussing, the exchange of women across borders actually facilitates, and stays within the model of, capitalism, merely seeking to (partially) extend its benefits to a slightly wider group. If we are inclined to see the movement of host-country women into the labor force, and their upward mobility within that labor force, as a good thing (either from the perspective of greater productivity or from a liberal feminist perspective), we could applaud the flexibility of transnational capitalism in easing the social costs of that transition.

• 24  The other part of this is a stereotype that women with children, or women who may at some point ha (...)
• 25  As economic growth proceeds (partly due to higher productivity of women in the professional workfo (...)

22But like many social issues, the issue of immigrant women “doing the dirty work” of cleaning, cooking and caring presents problems on various levels and for all governments and social groups involved. The feedback effects are key here. First, it maintains deep-seated social (and thus economic) inequities, because it relieves pressure on men – they don’t have to do housework because low-paid migrant women do their share and more. This reinforces the idea that is still pervasive in the US and much of Europe that housework is women’s work. Hiring immigrant women to do a family’s domestic work may ease marital relations, but at the expense of continuing stereotypes that poor migrant women can only do jobs based on “innate” feminine skills.24 Second, although the remittances paid back to the sending countries may help those countries, they do so at the expense of the care deficit Parreñas talks about with respect to the Philippines. Third, while we want to emphasize that it is restrictive policy, rather than lack of enforcement that is the problem, the current situation of migrant domestic labor does presenta policy problem for receiving governments, in that it undermines democratic values by permanently/essentially disadvantaging certain groups, and encouraging specifically gender-inflected forms of racism.25 The concern here is not so much that inflows of workers don’t match the levels officially desired by the receiving country, but rather that this institutionalized policy gap prevents acknowledging immigrants, and especially women who migrate to do domestic work, as “real workers” making an essential contribution to the economic health of the receiving countries. Such misrecognition contributes to the cycle of xenophobia and backlash in periods of perceived economic threat and does nothing to address underlying and persistent issues of gender inequity.

• 26  Functionalist arguments about immigration, such as the very public debate in the UK (summer 2008) (...)

23There is something ethically and politically troubling about the exploitation of women to serve the reproductive needs of capitalism, even if that exploitation can be “outsourced.” It seems that keeping borders formally “closed” when they are to all practical purposes actually rather porous is a quasi-deliberate strategy to maintain this low-cost service to capitalism. What underlies this is a pervasive functionalist approach to women’s position within society, which (as Susan Moller Okin [1989] has explained) asks what women are for, rather than seeing them as full social and ethical agents in their own right. The fact that some women escape this, on the backs of others, and that some (from both groups) may even benefit from it, does not remove this problem.26

• 27 Much fuller discussion of this is needed, but it seems essential to consider and understand refugee (...)

24From the supply side, it does not seem either practical, or ethically acceptable to discourage women from migrating to do this work. Plus, as Keely (2000: 58) has argued, the assumption that economic migration is voluntary needs to be problematized27.

• 28  See Williams and Gavanas (2008). In a three-country study, the authors found significant differenc (...)
• 29  See for example Anderson (2000), and compare Rollins (1985), Romero (2002), and Glenn (1986), on t (...)

26It is especially clear that the problem cannot be solved through calls for individual voluntarism, for instance by feminist appeals to middle-class women to treat their domestic workers “better.”28 Except in the case of the worst abuses, it is not even clear what this would mean: some of the ethnographically-based work, which has the great merit of giving voice to the women workers themselves, calls for increased professionalization of domestic labor, with attention (and compensation) based on tasks and schedules. But other ethnographies point to the dangers, and the abusive nature, of viewing care-giving as simply another sector of productive labor; some domestic workers complain of being treated like machines, or laboring animals. A complex ethical and social problem arises when a woman who has had primary responsibility for raising (and loving) a small child is fired and told she may never see that child again.29

• 30  See Anderson (2000).

27Ethnographies and narratives of all sorts emphasize the hollowness of assertions that the domestic worker is being treated “like one of the family.”30 At the first level of analysis, the problem is that this is a lie, but at a deeper level, the problem is the family, and the unique kinds of power differences, manipulations, pressures, and so forth, that are naturalized there and made acceptable by the continuing fiction of a separate or different “private sphere.”So insofar as the worker is seen as, or sees herself as, a family member embedded in a network of affective relationships rather than a working person with a set of contractual and/or inherent rights, her situation is both problematically different from and problematically similar to the women who are “real” family members.

• 31  For problems with “care,” see Narayan (1995).

28It seems important to recognize the “emotion work” involved in caregiving, and, on some level, perhaps even in “homemaking,” embedded as activities like cooking and cleaning tend to be in relationships which not only mimic, but also are, familial in nature. We can agree that domestic workers would be better off if they were paid more reasonable wages, and we can understand that continuing stereotypes about what is, and isn’t, “real work” – stereotypes which now have feminist versions – are part of what keeps wages low and conditions dire. But we cannot simply think ourselves out of this, on an individual basis. We do not disagree at all that men should share, or that the houses of the middle-class do not need to be as large, as luxuriously maintained, or even as clean and tidy, as upgraded standards of consumption now apparently dictate. But by itself this will not address the flows of global capital and labor, any more than walking to work will solve the problem of global warming. Parreñas’s label of “the international division of reproductive labor” strikes us as a more satisfying term than “global chains of care,” because it is less sentimental and more materialist.31

• 32  See for example Steedman (1987).

30Analysis of women’s work in migration runs up against some of the same theoretical tangles in conceptualizing women’s work that feminists have been looking at since the 1970s. Where is the boundary between paid and unpaid labor? Do women have “class” in the same way men do, is it more malleable, is it a helpful word at all?32 Are there some forms of purposeful activity that are not work; if so, how should they be compensated, and what should they be called? Does it even make sense to speak of domestic labor as a category? A job description that includes both skilled care of the elderly and washing the paws of the household dog twice a day is unlikely to be a useful tool of economic analysis. What ties the category together is not the type of work involved, but the fact that women (paid or unpaid) are the ones who do it, who are expected to do it. The regrettable fact that information on “informal sector” and undocumented work by women is barely being gathered may reflect a sense that this work is either “natural,” or devalued, or both; but it is hard to see how such information could really be gathered accurately, except in small-scale qualitative analyses. For instance, there is no way of counting the women (migrants or non-migrants) who stay in bad marriages, or “non-legitimized” dysfunctional heterosexual relationships, on more or less the same terms and for more or less the same reasons as other women migrate to do domestic labor for someone else.

• 33  For general critiques of the “human capital” approach, see Ferber and Nelson (1993). And yet, the (...)
• 34  See also Lutz (2008:3-4) “Migrant women are currently more educated than their predecessors; a sec (...)

31One thing does seem clear: a “human capital” approach is particularly unsuited to this problem. And perhaps it is unsuited to any analysis of “women’s work.”33 Because one of the particular features of being a woman in modernity is suppressing, or at least setting aside, the human capital that we have. Strober and Chan (1999) found this to be true for the elite graduates of Stanford and Tokyo University; it is also true for educated women from the Philippines, Latin America, or the former socialist republics, who find they can make more money as domestic workers in prosperous countries than by using their skills, and who choose to do so in large numbers, reminding us that “class” is a terribly slippery thing to apply to women. Once one accepts a certain crude gender configuration, “women” can always be downwardly mobile in ways “men” can’t – or perhaps won’t: there isn’t always construction work available for law graduates, but there are always babies who need to be diapered and toilets that need to be scrubbed; moreover, any woman can (at least in theory) swallow her pride (or maximize her utility) by cleaning houses or doing sex work. The finding in Morrison, et al. (2008: 5) that “Schooling is positively associated with international migration of females (to non-agricultural jobs), but not of males,” is suggestive in this regard.34

• 35  That it may be philosophically preferable to speak of women as agents rather than victims does not (...)

33As we mentioned earlier, it is problematic to try to separate “refugees” from “economic migrants,” both because conditions of structural violence include economic factors, and (more pragmatically) because people’s motivations may tend to be mixed (see Barile et al., 1994). Mixed motivations also can characterize some of the “choices” made by employers. It is problematic from the point of view of workers’ rights to see domestic labor as a “consumer good” rather than real work done by real people; but an analysis that sees it as also part of household consumption is not precisely wrong. What we need is a construction of “agency” that is more flexible and multi-faceted than the rational choice of homo economicus, who was neither a domestic woman migrant worker in the twenty-first center, nor the mother who employs her.35