‘ON PAR’? THE ROLE OF THE AU PAIR IN SWITZERLAND AND FRANCE

The role of the au pair as outlined in the 1969 Council of Europe's European Agreement on ‘Au Pair’ Placement is ambiguous. Although they provide child care for pay, au pairs are supposed to be treated as members of the host family. Based on in-depth interviews with au pairs in Switzerland and France, this paper examines how au pairs define their role. Results show how au pairs’ definitions of their roles vary and are constructed through interactions within the host family. The au pair's definition of her role is constructed through work and non-work dyadic interactions with the host mother. The relationship between mother and the au pair does not happen in isolation, however but is shaped by the intimate relationship between the host mother and host father. The triadic relationship between the au pair, host mother, and host father further shapes how the au pair defines and experiences her role.

In 1969, the Council of Europe drafted the European Agreement on ‘Au Pair’ Placement. France and Switzerland have since both signed the agreement. At its very conception, the au pair program in Europe was intended, at least on the surface, to be a cultural exchange for young people – not a source of employment. As defined by the Council of Europe, the au pair exchange provides an opportunity for young foreigners to visit a new country, learn the language and the culture, while temporarily living with and providing some services for a host family. While providing child care and light housework, au pairs are not coming to work but ‘share the life of the receiving family’. Au pairs do not have an employer but ‘host family’. They do not work for but rather engage in ‘day-to-day family duties’ for which they do not receive wages but ‘pocket money’. Indeed as the name au pair states, au pairs should be ‘on par’ or an equal member of the host family (Yodanis and Lauer 2005). Thus the role, or expectations for behavior, of au pairs is full of contradictions. Au pairs are not simply child caregivers. Rather, they are paid workers who are not employees. They are family members who are not members of the family. In this paper, we study how au pairs define their role.

The au pair exchange is an integral part of the global exchange of workers to and from Europe intended to meet the growing demand for low cost care providers (Yodanis and Lauer 2005). Research has documented the often structurally unequal status of immigrant care workers in relation to their employers. Pratt (1997), Anderson (2000), Hondagneu-Sotelo (2001), and Hess (2005), for example, describe the global migration of women from poorer to wealthier countries to work as caregivers for middle class families. These studies outline the impact of dramatically unequal ethnic and class statuses between employer and employee on employees’ work and living conditions. As cited in all of these studies, live-in arrangements, where employees’ work and living space are the same, and insecure immigration status leave women particularly vulnerable to and likely to be impacted by exploitation.

Like other caregivers studied, au pairs are foreign women who travel to a new country to provide live-in child care for middle and upper class families. There is a likelihood of their being exploited. Hess (2005), in her study of au pairs migrating from Eastern Europe to work as au pairs for Germany families, documents many of the vulnerabilities and the exploitation these younger women experience. In the most extreme form, there are cases of women being drawn and trapped into domestic slavery through au pair placements (European Parliamentary Committee on Equal Opportunities for Women and Men 2001).

Yet, au pairs are often different from other immigrant care workers in significant ways. While not always, au pairs can be ‘class peers’ of the host family (Wrigley 1995). Au pairs are often well educated and come from other middle class families and other highly developed, wealthy countries in western Europe or North America to work in a new country for one or two years for the cultural experience rather than economic necessity. Indeed, regulations have specified that au pairs must come from certain countries, generally those of similar wealth and culture, have completed secondary education, and speak the language of the hosting country. Yet even among au pairs who are ‘class peers’, some define their role as an on par member of the family, while others experience their roles as an unequal employee or servant. What explains the differences?

The European Agreement on Au pair Placement focuses on objective aspects of the au pair arrangements, such as working hours, pay, and living arrangements. These components of the au pair placement are clearly necessary for a non-exploitative relationship between au pair and host family. Yet as this study shows, these are far from sufficient components for ensuring that the au pair has the role of an ‘on par’, temporary member of the host family. This study finds that subjective aspects of the relationship between the au pair and members of the host family are central to whether au pairs perceive their role as either a member of the family or a servant.

Amidst discussions of the future of welfare states, increasing labor force participation among mothers of pre-school age children, and declining fertility rates, research on paid child caregivers has blossomed. Some studies have analyzed the interpersonal relationships between parents, children, and in-home childcare givers (Gregson and Lowe 1994; Macdonald 1998; Murray 1998; Wrigley 1995, 1999). Results show that the role of the caregivers within these relationships is ambivalent. On one hand, the relationship is described as ‘like family’. Caregivers express love and warmth for the children and treat them as if they are their own. Likewise, parents express warm feelings toward the caregiver, describing them as like members of the family. On the other hand, both caregivers and parents limit their feelings because of the economic, employment nature of the arrangements and efforts to maintain boundaries between the paid caregiver and the ‘true’ biological parents.

These emotional boundaries, or ‘detached attachment’ and ‘feeling rules’, can lead to an emotional dilemma among caregivers (Nelson 1990; Wrigley 1995). Au pairs may not like their work conditions or employers, but they can grow quite attached to the children (Hess and Puckhaber 2004; Hess 2005). Caregivers are expected and often want to invest their emotions in the relationship with the children. If they do not, their jobs can be alienating and unrewarding. But if they do, they are hurt when they no longer provide care after children grow up or if they are fired (Wrigley 1995).

Emerging from this ambiguity in roles, research has documented problematic dynamics in the relationships between caregivers and family members, including exploitation of the caregiver, conflict in the negotiation of boundaries, and mistrust over the quality of and disagreements over the best approach to care (Gregson and Lowe 1994; Wrigley 1995, 1999; Macdonald 1998; Uttal and Tuominen 1999; Anderson 2000; Hondagneu-Sotelo 2001). Role ambiguity is likely to be particularly true for au pairs as a result of the contradictions inherent in the au pair placement policies which state that au pairs, while working for and getting paid by the host family, are not employees but on a cultural exchange living as a temporary member of a host family.

From an interactionist perspective, roles are not fixed or static. Rather, roles are continually negotiated and emerge through social interaction with others. During these interactions, individuals take on and perform their roles to others in a social context, and by doing so, continually produce and reproduce roles (Turner 1956; Goffman 1959).

Also during interaction, some roles can be defined as either inferior or superior to others. Goffman (1956) outlined how inferior and superior status positions are created during interaction when through ritualized behavior some individuals display deference to others. Rollins (1985) applied this theoretical perspective to study the symbolic interaction between white female employers and African American female domestic workers that creates an unequal status between them, such as the employer being addressed as ‘Mrs.’ while the employee is called by her first name. Similarly, Macdonald (1998) shows how employed mothers and paid caregivers both use symbolic acts to ensure that the biological mother maintains the dominant caregiving role. While the paid caregiver may perform more of the total child care, certain tasks such as bathing or feeding are maintained as ‘mother only tasks’ and symbolize the biological mother's role as primary caregiver.

The primarily focus of research has been on the relationship between the mother and the paid caregiver, largely because the mother usually has the most contact with the caregiver (Macdonald 1997; Hess and Puckhaber 2004). Nonetheless the father or male partner may also have a significant impact on the work and experience of the caregiver. Extensive research documents the impact that spousal relationships have on other family members. Children can experience a wide range of behavioral and emotional problems, including aggression, anxiety or withdrawal, as direct or indirect consequences of conflict or violence between their parents (Katz and Gottman 1993; Wolak and Finkelhor 1998). While we know that the quality of spousal relationship affects others in the family, studies on paid caregivers have largely overlooked the impact of the male partner and the spousal relationship on the experience of caregivers living in the family.

In this paper, we examine how from the perspectives of the au pairs, their role as either a member of the family or employee of the family is constructed during daily interaction. In our analysis of interactions, we make a distinction between dyadic interactions between the host mother and au pair, and triadic interactions, involving the host mother, host father, and au pair. Following the classic work of Simmel (1950[1908]), qualitatively different interactions and thus consequences of the interactions are expected depending on the number of people involved. Simmel (1950[1908]) highlights processes possible within triads which are not possible in dyads. For example in triads, alliances can be formed between two participants versus the third. In other triads, the third person can work to divide and then dominate the other two participants. In yet other cases, the third person can take the role as a mediator between the other two participants. This paper examines how, from the perspectives of the au pairs, both their dyadic relationship with the mother and being embedded in a triad with the host mother and father shape their role.

The research presented in this paper is based on data from in-depth interviews with au pairs working in Switzerland and France. These two countries were chosen because even though they are geographically and culturally close, they differ significantly in their childcare policies, practices and norms.

According to the model developed by Gornick et al. (1997), France, along with Denmark and Sweden, has some of the strongest state-sponsored childcare policies, including financial support for families and child care benefits and availability. France has a universal system of family allowance. For instance, the «Allocation de garde d'enfant à domicile» (AGED) is given for each child younger than six years. The amount of this allowance ranges from a maximum of 1,500 euros to a minimum of 500 euros per trimester, depending on the age of the child and on the parents’ economic resources. Moreover, the ‘Aide à la famille pour l'emploi d'une assistante maternelle agrée’ (AFEAMA) supports parents who hire a person to take care of their children. In addition to these public measures to support families, the French government is also engaging in efforts to provide more grants for private childcare. According to statistics listed by Jenson and Sineau (2001), in France 57 percent of children aged zero and three years are cared for privately. Among the different private care providers, au pairs are well represented in France.

Switzerland follows familialistic gender and welfare regimes. This means that the welfare system is based on the assumption that the care of family members, including children, is the responsibility of biological family members, particularly mothers (Drew 1998). As a result, the state does not provide universal, public childcare services (Dafflon 2003). If the family is not able to provide care, this must be purchased on the market (Esping-Anderson 1999). Yet private care is also limited and very expensive (Charles and Buchmann 1994). So, in the Swiss context, au pairs are likely to be one of the least expensive options, with a required minimum monthly wage of 700 CHF plus room, board, and tuition for language courses (Mellini et al.2005).

For this study, au pairs were recruited through advertisements in au pair placement agencies and on websites, as well through snowball sampling techniques. To homogenize the sample and ensure sufficient experience as an au pair, au pairs had to be female and working for at least one month to qualify for the study. The final sample includes 20 au pairs from 12 different countries, including countries in Western Europe, Eastern Europe, North America, and South America. In accordance with the policies, the au pairs range in age from 18 to 27. For all but two, it is their first experience as an au pair.

Data were gathered through semi-structured interviews, which lasted about one and a half hours and were taped and transcribed. The au pairs chose the language of the interview, since the interviewer is fluent in both French and English. Half of the au pairs chose to conduct their interviews in French and half in English.

Data gathering and analysis were conducted in three stages. Each stage consisted of interviewing some of the au pairs and conducting preliminary analysis, which guided data gathering in the subsequent stage. Data were collected through the biographical accounts method, in which respondents do not necessarily provide a complete life history but detail about certain stages in their lives (Ricoeur 1983; Demazière and Dubar 1997). During semi-structured interviews, au pairs were asked to describe their lives just prior to and during their au pair experience. They were asked about a number of topics, including their reasons for becoming au pairs, their expectations of the work, their actual work situations, every day life as au pairs, and their social networks. While discussing the topics proposed by researchers during the interview, the au pairs were able to assemble and integrate an array of seemingly unconnected events of their experiences into a unique, coherent account that is meaningful to them.

Thus, the data are from the perspectives of the au pairs. There are clearly limitations to this data, given the focus on interactions within dyads and triads. A systematic and diachronic approach to data gathering would be helpful for gathering information from the viewpoints of others involved in the relationship and how those perspectives change over time. Observational or ethnographic research would also add a different perspective to the interaction. Yet it is the au pair whose perception and experience is the focus of this research.

Assisted by NUD.IST qualitative data analysis software, analysis was conducted using a grounded theory approach and constant comparative method (Glaser and Strauss 1967; Ragin 1987; Demazière and Dubar 1997). During this process, a two-staged thematic analysis was conducted (de Sardan 1995). The first analysis involved highlighting particular topics suggested and discussed by interviewees in relation to the original research questions of the study. The second analysis, a cross-interview analysis, completed the first by considering evidence from the interviews that either confirmed or refuted and altered emerging patterns.

In this section, we discuss how the role of au pairs varies and emerges from interactions with members of the host family. The au pair experience and relationship with the host family is complicated, shaped by many factors including but not limited to similarities and differences in the cultural and life histories of the au pair and all members of the host family and prior expectations and preconceptions of the relationship and each other, often based on previous experience working as or hiring an au pair (Mellini et al.2005). Keeping in mind these complexities, we focus in this paper on what transpired from the interviews as the most significant factors affecting the au pair experience, as perceived by the au pairs. First we discuss how the au pair's definition of her role is constructed through subjective work and non-work dyadic interactions with the host mother. Second we explain how the triadic relationships between the host mother, host father, and au pair further shape how au pairs define their role.

Au pairs define their role in different ways. Some view themselves as ‘on par’, members of the host family, as outlined by the European Agreement on Au pair Placement. For example, one au pair explains, ‘I feel treated like a family member’. Likewise, another au pair explains, ‘I don't feel I have a relationship where they're my boss and I'm their employee … I am considered a part of the family’. In comparison, some au pairs define their role as ‘simply an employee’, a ‘servant’, or even a ‘slave’ within the family. As an au pair says of her experience, ‘I was feeling like a slave’. Another au pair says, ‘I was a servant’.

What explains these different experiences? In this section, we argue that the au pair's role as either a family member or employee is shaped and defined through subjective work and non-work dyadic interactions with members of the host family.

Different from Hess's sample (2005; Hess and Puckhaber 2004), the au pairs in our sample tend to be ‘class peers’ of the host families for whom they work. All have at least a high school degree. In addition, four au pairs have a bachelor's degree and one has a master's degree. All but three of the au pairs come from wealthy, western European countries and middle class families. The four au pairs who come from less wealthy countries in Latin America or Eastern Europe, come from wealthier families within those countries. Indeed the one au pair from Latin America felt uncomfortable in her work because she was used to having servants herself at home. Another good example of the ‘class peer’ status of au pairs is that two of the host mothers themselves worked as au pairs when they were younger.

While there is variation between au pairs in pay, working hours, vacation time, and job requirements, including the balance of child care to housework, none of the au pairs in our sample were working under highly exploitative condition. Pocket money ranges from approximately 60 to nearly 200 Euro per week. In Switzerland working hours range from 30 to 40 hours per week (six hours per day, with two days off), depending on cantonal regulations. Au pairs also generally have four weeks of vacation (five weeks until the age of 20). In France, working hours are fixed at 30 hours (five to six hours per day), plus two or three evenings of baby-sitting per week. Au pairs in France have the right to two weeks of vacation.

It is important to note that while there is variation among au pairs in these objective conditions of work and while there are significant differences between France and Switzerland in au pair policies, the variation does not fully correspond with or explain the construction of the au pair's status as equal or unequal to members of the host family. While some au pairs were frustrated with some of these objective aspects of their work, none cited them as a primary reason in the construction of their status as equal or unequal to the host family. For example, one au pair works long hours and another has no paid vacation but they both are happy and feel treated like family members by their host families. In contrast, another au pair who feels treated like a ‘slave’ is the highest paid au pair in the sample. And while she works long hours, she emphasizes that the hours are not what bothers her, ‘I never complained about work. That's just the way I am … I don't even make any faces … I don't mind really working. That's not a problem for me’.

Furthermore, there is more variation in au pairs’ roles and experiences within Switzerland and France than there is between the countries. Regardless of the differences between the countries in childcare and au pair policies, some au pairs within the both countries define their roles as an ‘on par’ member of the family, while others see their roles as that of an unequal employee or servant.

Symbolic work and non-work interactions emerge as particularly important for the au pairs’ construction of their status as equal or unequal. The dyadic relationship between the au pair and host mother is most significant because it is usually the mother who is most actively engaged with the au pair (Wrigley 1995, 1999; Hess and Puckhaber 2004; Hess 2005). During interactions, the host mother can either demonstrate to the au pair that she is a member of the family and thus equal to her, or show her that she is merely an employee or servant, and thus unequal.

Overall, the au pair is defined as an equal member of the family when the au pair feels that the host mother is defining the au pair's role as similar to her own. For example, the role of the au pair is constructed as an equal member of the family when the au pair, not only the mother, is trusted to set work tasks and schedules. In addition, consistency between the host mother and au pair is important. Equality is demonstrated to and constructed for the au pair if the host parents themselves follow the childcare and housework guidelines they establish. The au pair defines her role as an employee of the family when the parents keep different standards for themselves. For example, one au pair's host parents ask her to clean up the kitchen after she uses it, while they never do it themselves. Another au pair is asked to cook healthy food, while her host parents just cook pasta. Still other host families ask the au pair to limit the amount of television the children watch but then when they are home, they let the children watch TV whenever they want.

Reciprocity in performing and receiving favors also plays an important role in constructing equal relationships. Equality in the relationship is perceived by the au pair when a fair exchange occurs between host mother and au pair. In other words, a more balanced give and take in a relationship is symbolic of equals – whether family members or friends. An imbalanced relationship with one participant giving and the other taking is characteristic of an unequal, servant relationship (Gouldner 1960; Rosenbluth et al.1998).

If the au pairs have to do menial tasks for the parents, in addition to the kids, they perceive the expectations as unfair and are made to feel unequal.

Doing tasks for the parents does not have the same effect, however, if the parents reciprocate and do things for the au pair.These reciprocal exchanges are symbolic of not only fairness and equality but also family exchanges. They are the types of interactions that family members should do for each other as part of their ‘day-to-day family duties’.Even the simple act of being recognized and thanked for their work makes the au pairs feel like their work is valued and not being taken for granted (Hess 2005). This again is symbolic of family relationships where appreciation rather than money is exchanged for helping one another.

Thus when the roles of the host mother and au pair are defined by the au pair as similar in relation to the work that needs to be done in the family, the au pair defines her role as equal to rather than serving the family.

Non-work interactions can be even more important than work interactions for constructing the au pair role. Overall, au pairs define their role as an equal member of the family when their interactions with the family are not merely work-related but extend into non-work time. This is particularly significant given that both the au pair's work and living space is within the host family's home. Equality between the au pair and host family members is constructed when au pairs are made to feel comfortable using the house for non-work activities, including watching TV, the internet, telephone, and entertaining friends and when au pairs are invited to participate in family events and celebrations.

Again these interactions are all basic to family relationships and thus, symbolize to the au pair that she is equal to members of the family. On the other hand, if the au pair is made to feel that the host family is discouraging any non-work interactions, she feels inferior and defines her role as an employee to the family.The host family can place limitations on many aspects of the au pairs’ daily life activities, by accusing them of eating too much or prohibiting them from eating the food that is available or restricting how much they shower or how much hot water they use (Hess 2005).

Conversation is another important aspect of non-work interaction. In particular, an equal exchange between au pair and host mother of personal information which is not about child care or housework constructs an equal relationship. When the mother discloses information about her private life to the au pair while also asking the au pair about her life, including her home country and family, au pairs report that a peer rather than an employer–employee relationship develops.

In turn au pairs feel equal when the host parents reciprocate by sharing information about themselves. This is particularly true if host family members disclose vulnerabilities, such as a host mother revealing her own doubts about parenting:In other cases, feelings of homesickness can be shared when the host mother, not only the au pair, recently moved to a new culture. As one au pair explains, ‘We speak a lot, especially with [the host mother], a lot about homesickness, because it's quite similar to her feelings. She's homesick as well, you know, she misses friends and all her family. And so on. So, we speak a lot about this’. Reciprocity in sharing personal information is central to the establishment of an equal relationship. As Rollins (1985) found, if personal information only flows one direction – from the employee to the employer – the employee is made to feel vulnerable and used. Similarly, however, when host parents do not bother to ask au pairs about themselves, au pairs receive the message that they are not interested in them as people, only as employees.

Thus, through symbolic displays during work and non-work dyadic interactions between the au pair and mother, the au pairs come to construct their role as either a member or employee of the family.

The dyadic relationship between mother and caregiver does not happen in isolation, however but rather happens within the context of the other family relationships, most significantly the marital relationship between the mother and her spouse. In this section, we discuss how the triad of the au pair, host mother, and host father further shapes how the au pair defines and experiences her role in the family.

Close relationships refer to spouses who are physically present, communicative, and compatible. An au pair describes this type of relationship between her host parents:

The easiest situation for the au pair tends to be when the relationship between the father and mother is good. Within this context, the relationship between the mother and the au pair tends to be equal and close and it is also possible for the au pair to have a comfortable and close relationship with the father.Here the au pair is talking about the positive and equal relationship she has with both parents, which in our sample is quite common, although not universal, within the context of close relationships.

Distant relationships refer to spouses who are either physically separate, as a result of divorce or work demands, or are emotionally or communicatively detached. Another au pair describes a distant spousal relationship in her host family:

When only the mother is around, whether as a result of the father being physically or emotionally distant from the family, the context may become more complicated for the au pair. When only the mother is present, the mother–au pair relationship is central to her experience. When her role is defined as a member of the family, the au pair may further define her role as helping the mother who is alone. This is similar to the triad-based alliance outlined by Simmel (1950[1908]).When the father is rarely present, the au pair may feel awkward or uncomfortable interacting with him when he is there. This, however, does not seem to have a strong negative impact on the au pair's role or experience.The real downside of this context comes when the au pair role is defined as an unequal, employee or servant of the family. When the au pair feels treated by the mother as inferior and the father is not around, there is no other adult person in the house to witness and buffer potentially negative treatment, either directly or indirectly. In other words, there is no third person present in the ‘mediator’ role suggested by Simmel (1950[1908]). These are some of the worst relationships for au pairs and the conditions under which the au pairs are most likely to leave. One au pair who ended up leaving her job described her experience as follows:This treatment by the host mother had serious consequences for the au pair.

Abusive relationships refer to spouses who are engaged in extended conflicts or relationships characterized by verbal, emotional, or physical abuse. One au pair describes such a relationship between her host parents:

The context of a conflict-ridden or abusive relationship between the father and mother also impacts the au pair's relationship with the mother and her role in the family. While there are only three cases of abusive spousal relationships in our sample, the examples are illustrative of the potential consequences of these relationships. First in the context of abusive spouse relationships, the au pairs in our sample never report a good relationship with the father. Indeed at times, they were themselves targets of abuse by the fathers.

The au pairs’ role in relation to the mother can also be shaped by the unhealthy relationship between the mother and father. In particular, au pairs who are treated by the mother as an equal member of the family may form an alliance with the mother against the abusive father, similar to what happens when the mother is alone. These au pairs tend to feel a sense of loyalty toward the host mother, whom they define as a victim, and feel an expectation to help her. Although they originally took the job to have the experience of living in a new country while caring for children, an au pair in this context can reconstruct her role into an advocate for the mother. They perceive their work as someone whom the mother can talk to about the problems in her relationship and they provide advice on leaving the relationship.

Their protection of the mother can also lead to support for the mother even as she uses child rearing practices that they see as unreasonable. Au pairs in these contexts, while not having the positive cultural experience they expected, tend not to leave their position out of loyalty for mother.Later this same au pair describes the close, supportive relationship she has with the mother, ‘I get along really well with the mother. We talk all the time. She said, “I know your door's open to me, and mine is open to you”’.

If the au pair feels treated as an unequal employee by the mother, she may not so clearly take on the role of supporting the mother as an ally against the father. There is only one case of this in our sample, but it is quite illustrative. Instead of being allies against the father, the abuse can result in the au pair feeling divided from the mother. This is similar to the divide and dominate dynamic outlined by Simmel (1950[1908]). In the following example, the au pair describes the mother being somewhat supportive of her, but also taking the father's side against her.

Similarly, as she describes in her reaction to seeing the father fight with the mother, the au pair feels ‘good’ that she is not the only one being treated badly by the father.Still, the au pair blames the father rather than the mother for the problems in her relationship with the mother.And even though the au pair does not take on the overt role as an advocate for the mother, she still realizes the expectation that she could reconstruct her work into helping the mother.As this section outlines, the au pairs’ definitions of their roles are also shaped by the fact that the relationship between the mother and au pair is embedded within a triad, including the host father. The quality of the relationship between the mother and father further shapes the interactions between the au pair and host mother and the au pair's perception of her role within the family.

This paper examines how au pairs’ definitions of their roles vary and are constructed through interactions within the host family. During dyadic interactions with host mothers, work and non-work interactions can be symbolic for the au pair of either equality and family relationships or inequality and employee relationships. The au pair's definition of her role is shaped by these interactions. In addition, the interactions between the host mother and au pair are embedded in relationship between the host mother and host father and these triadic interactions further shape how the au pairs define their role.

This study finds that objective aspects of the au pair arrangement, such as working hours, pay, and living arrangements, are not sufficient for shaping an au pair's perception of her experience and role. Rather subjective aspects of the relationship between the au pair and members of host, such as reciprocity in doing favors sharing information, and engaging in family activities, matter for the au pair's role. In addition, whether the host father and mother are close, distant, or abusive shapes how the au pairs define their role within the family. These aspects of the experience have real consequences for how au pairs live and define their experience.

These findings are also relevant to and indeed consistent with Hess's (2005) study focusing on au pairs who are in unequal structural positions, backed by racism and classism, in relation to their host family. The racism and classism the au pairs experience is not only about national policies and socioeconomic differences, rather the inequalities take form and have consequences through the daily, symbolic interactions with the host family (Rollins 1987).

In considering these results in the context of caregiving research, it is important to again emphasize that au pairs are unquestionably a unique form of caregiver, which impacts the findings. For example, au pairs, as a requirement of the job, share a home with the family and by definition should be treated like a member of the family rather than an employee. These characteristics should increase the impact that the quality of other family relationships has on the experiences of au pairs.

In addition, the fact that au pairs, more than other caregivers, tend to be ‘class peers’ of the families for whom they provide care may make reciprocity in work and non work interactions as well as the au pair's role as an ally to the mother facing an abusive husband uniquely possible. For example, conversations with the au pair can be more comfortable for the mother, because au pairs may discuss problems related to homesickness or boyfriends but ‘seldom bring the parents face to face with the hardships experienced by [the] poor’, as other domestic workers might (Wrigley 1995: 170).

In sum, the results presented here are admittedly preliminary and exploratory. Based on 20 interviews limited to the perspective of au pairs, of which only three involved highly negative spousal relationships, the findings should be viewed cautiously. However, the data do point to the need to consider new directions, previously overlooked, in future research and policies related to au pairs in particular and care work, more generally.

As increasing numbers of families solve challenges in combining work and family by bringing outside workers into the family, a corresponding pool of research is examining these unique work/family relationships. Existing research, however, tends to draw more from the work rather than family literature. This study suggests that a true blending of work and family research topics may be necessary for understanding these complicated work conditions. Issues central to the study of families, such as spousal violence for example, are likely not separate from but integral to the work experiences of in-home paid caregivers.

This research was supported with a grant from the Swiss State Secretariat for Education and Research.

Laura Mellini is a sociologist affiliated with the Department of Social Sciences, University of Fribourg (Switzerland). Her main interests are on HIV/AIDS, homosexuality, family and identity construction.

Carrie Yodanis is s sociologist and assistant professor of family studies at the University of British Columbia. She also conducts cross-national research in the areas of gender inequality in intimate relationships and violence against women.

Alberto Godenzi is Dean and Professor in the Graduate School of Social Work at Boston College His research interests include interpersonal violence, peace and conflict studies, HIV/AIDS, international social development, higher education.