Skip to content

About the au pair project

  • The aim of this two-year ESRC funded project is to provide evidence about the experiences of au pairs and host families in the UK since the ending of the au pair visa in November 2008. This change, which was part of the move to the UKs Points Based Immigration Scheme, is important because it reduced government control of the au pair scheme and there is evidence that numbers of au pairs have been growing in a decreasingly regulated environment. Au pairs are now one of the cheapest and least regulated forms of in-home labour and are relied on by many thousands of households, yet little is known about au pairs in the UK or the people who host them. For many families in the UK, and elsewhere in Europe, the solution to the problem of balancing work and home life is to rely on the low waged labour of migrant women to perform domestic labour.
  • This project therefore, brings together two important issues for contemporary society – women’s changing relationship to the home and the growth in labour migration. Au pairs are particularly invisible to the official gaze because they are not technically ‘employed’ but occupy a liminal position between workers, students and working holidaymakers. Since the ending of the au pair visa there is effectively no official data gathered on au pairs in the UK. Our aim is to gather both quantitative data to describe recent trends within au pairing as well as qualitative data to understand the part that au pairing plays in the lives of au pairs and their hosts.

Podcast – ‘Au Pairing After the Au Pair Scheme’ – ESRC Research Project Dissemination Event’

The Birkbeck Institute for Social Research, Birkbeck, University of London presents:

‘Au Pairing After the Au Pair Scheme’ – ESRC Research Project Dissemination Event

Bridget Anderson (COMPAS University of Oxford)
Nicky Busch (Birkbeck)
Rosie Cox (Birkbeck)
Helle Stenum (Roskilde University, Denmark)

Au pairs on a pittance: the young women minding kids for £2 an hour… media coverage of the research project

Some of the findings of the au pairs project have been reported recently in the Guardian, The Times, The London Evening Standard and on Woman’s Hour.

Links below:

The Guardian:

The Times:

Woman’s Hour

The Evening Standard

Findings from review of migration and employment data

Measuring au pairs in the UK – quantitative data sources and their limits

Compared with the overall number of domestic workers in the EU, and more specifically in regions across the UK, the number of au pairs living in private homes can be assumed to be relatively small. However, arriving at a definitive and verifiable total number of persons employed as au pairs is not possible, particularly since the abolition of the au pair visa in 2008. This is because 1) the number of EU au pairs is not registered or measured accurately by any of the existing data sources and because 2) the concept of au pair is also used in various irregular forms of employment. Below, we outline first the available figures that can be used to gain an indicative account of the scale of au pairing in the UK pre-2008, when figures for au pair visas granted were available, before discussing alternative data sources and their deficiencies that can provide information relevant to the pre and post 2008 periods.

Measuring au pairing in the UK pre-2008
In the years before 2008, the UK recognised three groups of European countries in relation to the allocation of au pair visas. As of December 2002 these were constituted as follows. First, individuals from the European Economic Area (EEA) did not require visas to work in the UK as au pairs. Second, nationals of the following countries required visas to work as au pairs in the UK: Andorra, Cyprus, the Czech Republic, Estonia, Faroe Islands, Greenland, Hungary, Latvia, Lithuania, Malta, Monaco, Poland, San Marino, and Slovenia. The third group required visas to enter the UK, as well as to work as au pairs: Bosnia, Bulgaria, Croatia, Macedonia, Romania, Slovakia, and Turkey.

Au pairs coming from EEA countries – how to measure in the absence of visa allocation?

The total number of people employed as au pairs in the UK was unknown, even in the era of the au pair visa, because many au pairs were from EEA countries, and did not require visas. This was confirmed by Cox’s (1999) survey of UK placement agencies. Her interviewees considered that the main source countries were France, Spain, Italy, and Germany. However, they also noted that “some Eastern European countries are becoming increasingly important sources of au pairs”.

Measuring au pairing from non-EEA source countries
Of those people entering the UK as an au pair who did require a visa, in 2001, au pair visas were issued to 12,000 people; in 2002 the figure was 12,800, in 2003 au pair visas were issued to 15,300 people; in 2004 it was down to 5,640; 2,360 in 2005, 1,840 in 2006 and 765 in 2007 (the figures after 2004 reflected the fact that accession country au pairs no longer needed visas). However, while the number of visas issued provides an indication of the numbers of au pairs coming from certain countries, it is likely that a very significant number of people working as au pairs would have entered under a different visa, not needed a visa (eg because they were from an EU state) or would have been working illegally (Cox 2006). The removal of a specific au pair visa in 2008 meant that after this point it was envisaged that all au pair vacancies would be filled by EU citizens, as well as by a very limited number of visa holders from Australia, New Zealand, Canada and Japan. The number of people working as au pairs who did not require a visa was not, however, recorded on any official data sources. Moreover, the number of people who had entered the UK and/or were working as au pairs irregularly is very difficult to estimate.

Other data sets

Other data sets can give an indication of the size of related or complementary labour market sectors, but no data source accurately measures the number of au pairs in the UK post 2008, which countries they have come from, how long they stay or whether they return to their home countries. Other data sets can also provide information on the scale and extent of new migration generally, the regional distribution of new migrants, and the make-up of the migrant population in terms of countries of origin, age, occupations and formal education levels. This more general information can be used as a background to the more nuanced and specific quantitative and qualitative research methods necessary to learn about the scale and distribution of au pairs in the UK.

The Labour Force Survey

The UK ONS Labour Force Survey, for example, measures domestic workers, but this does not include au pairs. The Labor Force Survey is a quarterly sample survey of 100,000 workers in 60,000 households. The LFS identifies domestic workers in its standard industrial classification series. The definition is as follows:

Code 97.00 Activities of Households as employers of domestic personnel
This class includes the activities of households as employers of domestic personnel such as maids, cooks, waiters, valets, butlers, laundresses, gardeners, gatekeepers, stable lads, chauffeurs, caretakers, governesses, babysitters, tutors, secretaries etc.

The LFS is likely to under-report domestic workers as a class. In addition, it tends not to report workers in the grey economy. Labour market intelligence from the TUC suggests a disproportionate number of domestic workers are employed in the grey economy.

In 2009, domestic workers numbered 52,000 (1.0 per cent of the employee workforce and 1.8 per cent of the total workforce). Female domestic workers make up 1.6 per cent of the employee workforce and 2.7 per cent of the total female workforce.
The survey also records that 10,000 people are domestic workers in their second job. This group is overwhelmingly female and London has the largest concentration of domestic workers – 18,000 (34.6 per cent).

The Labour Force Survey did indicate an increase in the employment of foreign-born workers in the ‘domestic sector’ in London, with the percentage of foreign-born workers employed in this sector across the UK rising from 9 per cent in the period 1992-1995 to 16 per cent in the period 2004-2007. In London, in the period 2004-2007, according to LFS data, 57 per cent of domestic workers were foreign born, compared with 45 per cent being foreign born in the period 1992-1995 (Kilkey and Perrons 2008).

Research by Kilkey and Perrons found that for the period 2004–07, while ‘foreign-born’ were over-represented among domestic-sector workers in the UK, constituting 16 per cent of such workers compared with 10 per cent of all workers, handyman-type work was less migrant dense than female areas of domesticwork; a finding suggest may be related to gendered constructions of ‘skill’, which in turn impact on pay and conditions of employment, and ultimately on the constitution of appropriate workforces. However, compared with the profile of the domestic-sector workforce some ten years previously, Kilkey and Perrons observed that ‘foreign-born’ men’s presence in handyman-type work has been increasing and has been doing so at a faster rate than UK-born men’s. They also demonstrate that ‘migrantisation’ of the domestic-sector workforce is greatest in London, where 57 per cent are ‘foreignborn’; this compares with 45 per cent some ten years previously.

In the period 2004-2007, 13 per cent of this domestic sector workforce was employed in childcare and related personal services. The data sources were insufficient for capturing self-employed and informally employed foreign-born workers providing childcare and related domestic services in London. However, it was clear that the domestic and childcare sectors in London were bolstered by the increasing number of female migrants entering the UK, particularly from the A8 states and especially in London. Note again, though, that while these figures provide an indication of the scale of the domestic work sector, au pairs are not included in these figures – because they are not held to be ‘employed’.

The Worker Registration Scheme

The Worker Registration Scheme was introduced on 1 May 2004. It required that nationals of the new EU member states (except Cyprus and Malta) who intended to work for more than one month for an employer in the United Kingdom were required to pay £70 and provide a letter that demonstrated that he or she had work arranged (Home Office 2004). After having worked legally in the United Kingdom for 12 months without a break, these workers were granted full rights of free movement and were no longer required to register on the Worker Registration Scheme. Instead, they could get a residence permit that confirmed their right to live and work in the United Kingdom. It should be noted, though, that a) there was significant under registration in the WRS; b) that au pairs were exempt from registration. While it is not clear that the WRS acted as a disincentive to citizens of A8 states wishing to enter the UK (perhaps because people are coming with the intention of working rather than seeking access to benefits), the scheme did provide a new source of data that could be used to inform attempt to understand migrant flows into the UK labour market. Again, however, while it gives an indication of stocks and flows of migrants from 2004 to 2011, when it was abolished, it does not measures au pairs as a distinct category.
The WRS shows that although the sharp increase in net migration flows to the UK predated EU expansion in 2004, the new arrivals entering the UK from the late 1990s were joined by 427,000 people registered for the Worker Registration Scheme between May 2004 and June 2006 (125,885 WRS applications were approved in 2004; 204,970 approved in 2005; 227,875 in 2006; 210,805 approved in 2007; 158,340 in 2008 and 108,650 in 2009 and 22,810 in Q1 2010). The largest number of people in every year came from Poland (between 2004Q2 and 2006Q3 over 300,000 migrants from Poland had registered on the scheme, which was 63 per cent of the total), followed by people from Lithuania (11 per cent) and Slovakia (10 per cent) (Home Office 2010).

The WRS data does go down to narrow occupational classifications, but it does not distinguish people working in private domestic settings from people working in childcare and commercial cleaning, for example. This is because the Standard Occupational Classification (SOC), on which the figures are mostly based, do not distinguish between different forms of household employment. The relevant SOC codes are 6122 (child minder and related) and 9233 (cleaners and domestics). In period April 2008-March 2009, there were only 145 migrants recorded as being employed in SOC codes 6122 and 8,790 in SOC code 9233. However, SOC code 9233 mixed domestic and commercial employment, including agency cleaning staff.

A further significant problem with using the WRS to measure the number of people working in private homes is that self-employed people were not required to register. Given the nature of employment in private homes, many foreign-born workers employed as childcare workers and cleaners, where they were in formal employment at all, would be ‘self-employed’ so they would not be recorded by WRS. On this point, the Annual Population Survey showed a sharp increase in female self-employment since 2004 (seven per cent compared to a two per cent rise in the female labour force). This pattern was different for female employment than it was for male employment (i.e the number of people who were self-employed rose faster than the number of employees).

There was also an increase in the number of women from A8 states entering the UK so that while the male to female ratio of WRS applicants in the period May 2004 to March 2009 was 56%:44% by Q1 2009 the figure was 50:50. It is also of note that the majority of applicants were young: in the 12 months to March 2009, 78% of registered workers were aged 18-34. And nearly 85 per cent of migrants had an educational qualification (Home Office 2009).

National Insurance applications
Natinal Insurance applications do not record job title, and as such do not provide evidence of numbers of au pairs specifically. However, this data source is useful for gaining insight into increases and regional dispersal of new migrants. National Insurance Number applications (up from 12,000 in 2002-2003 to 270,000 in 2005/06. Between April 2004 and March 2006, approximately 380,000 people from A8 states had registered for a national insurance number) (Blanchflower et al 2007). National Insurance Number allocations to EU accession country nationals show a jump from 10.35 (000s) in the first quarter of 2004 to 33.13 in the final quarter of 2004. NI allocations to A8 nationals peaked at 115.75 (000s) in the first quarter of 2007, before decreasing to 42.97 in the final quarter of 2008.

New arrivals from A8 states were not concentrated in London, but were instead spread around the UK with the East of England receiving the largest percentage (15 per cent or 73,035 people), followed by Greater London (13.1 per cent of 63,795 people) and the Midlands (12.3 per cent or 60,010), Central England (10.2 per cent or 49,740) and other regions receiving percentages under ten per cent. This suggests that workers from the A8 states were not drawn to the London labour market specifically but were instead employed in a range of industries and regions across the UK (Greene et al 2008).

Using data sources to measure migration and patterns of settlement – building a background picture
The LFS, the WRS and NI application figures are therefore insufficient for measuring the number of women from A8 states and elsewhere working as au pairs in the UK, and not robust for measuring the number of UK-born people and migrants working as nannies, mother’s helps, maids etc in households across the UK. It is clear, however, that there has been an increase in the total number of newly arrived migrant workers in London from the late 1990s, and again after 2004, and that a significant proportion of new entrants from A8 states in particular are female and relatively young. Data sources can therefore be used to build a background picture for measuring au pairing the UK that takes account of the extent of new migrations in specific regions and the make-up of migrants in terms of age, education levels, countries of origin etc. However, these figures must also be qualified by figures showing significant return migration, particularly in the wake of recession in the UK

A8 worker return migration and slower entry after 2008
UK government migration policy post 2008 was informed by the notion that A8 workers would fill all vacant low-paid jobs in the UK – therefore allowing for ever more restrictive migration policies being applied to non EU migrants. Following accession of the new EU member states the UK was the main destination country, receiving almost one-third of the estimated two million A8-nationals to have migrated to the old member states (EC, 2009b). A slight majority (56 per cent) of those registering with the Worker Registration Scheme (WRS) to work in the UK were men, and in terms of nationality, the vast majority (66%) were Polish (UK Border Agency, 2009).

However, past recessions affecting the UK saw a fall in the number of new migrants coming to the country to seek work, and this pattern was repeated in the wake of the recession that affected the UK economy from 2008. The Migration Advisory Committee (MAC) points out that in this period, the UK sent more people abroad to work than accepted new arrivals. There was a significant decline in the National Insurance Numbers allocated to non-UK nationals, particularly for “A8 nationals” (citizens of states who joined the European Union in 2004). A8 nationals fell by over 20 percent in the year up to September 2008. Data indicates in particular a significant falling off of A8 national migrants in low-wage sectors (and other sectors) in the first and second quarters of 2009. Importantly, however, this is telling us about flows but not stocks: people do not de-register, and it is estimated that about half of those registered have left the UK. The existence of significant return migration is further supported by Eurostat data suggesting that there were no significant declines in population in any of the A8 countries, especially Poland between 2004-2005 (Blanchflower et al 2007). In addition to significant return migration, it is clear that fewer people came to the UK from A8 states each year from 2007. This deos support the views advanced by the UK migration policy makers that EU accession state ‘migration’ would be particularly beneficial to the economy of the United Kingdom as it would effectively capture lower-skilled migration where it was needed to fill short-term or evolving gaps in the labour market, without associated costs of settlement, and that supply would reflect demand in the labour market as fewer people would come searching for work in times of recession.

Given the relatively short time that has elapsed from accession of the poorer EU states at the time of writing, as well as the need to account for the longer term effects of recession in the UK and elsewhere depressing the labour market, it not possible to make a definitive statement about the scale of EU accession state migration to the UK – and hence longer-term impact on labour markets across the UK. However, it appears that while the very large number of accession state workers, particularly Polish workers, who came to the UK immediately after 2004 was not sustainable, A8 workers remained a significant sector of labour markets in London and in other areas of the UK, with the potential that numbers could increase again in a more favourable economic climate.

In summary, no one ONS and or other UK government department data source is sufficient for measuring size and distribution of migrants working as au pairs in the UK. The number of people entering as au pairs was not accurately measured before 2008, and then not recorded at all after 2008 – and the data source that did measure employment of ‘domestics’ did not break down categories of workers so that it was possible to differentiate nannies from housekeepers, for example. Second, much domestic work was performed by workers behind closed doors in private homes, and the very ‘private’ nature of this environment meant it was not captured by official data sources (Lutz 2008). Third, the ‘female’ nature of childcare and housework meant it was often not regarded as not a ‘proper job’, and was thus treated as something that could be rewarded with pocket money or cash in hand payments (much as you would pay a babysitter who cared for children for an evening) (Ozegin and Hondagneu-Sotelo 2008). However, data on migration stocks, flows and regional distribution is useful for providing background to the population from which au pairs are drawn.

Findings from interviews with au pair hosts

We interviewed 15 au pair hosts who between them had hosted over 50 au pairs over a number of years. These interviews were extremely fruitful for discovering how some hosts understand and negotiate relationships with au pairs, how these change over time as children age and as hosts become more experienced at living with au pairs.

Aims of the interviews
We aimed to interview 40 people (men and women) who had employed or who currently did employ someone they identified as an ‘au pair’. We wished to learn more about why the services of an au pair had been required. Was it for cultural exchange? To have someone else in the house for security or for company? Was it a language exchange opportunity? Was it a way of getting someone to do childcare and housework for less money than it would cost to pay a cleaner and a nanny? We also wanted to know about employers’ childcare philosophies and how hiring an au pair sat with those. We were interested in how people went about hiring an au pair and how they experienced issues of intimacy, home and self when an au pair was present in the home. We wanted to know how hiring an au pair affected relationships between mother, father and child(ren) and between parents. We were also interested in whether employers felt there was an ethnic hierarchy at work when they hired an au pair. Did employers have preconceptions about what nationalities made the best au pairs? Finally we wanted to know the answers to more prosaic questions such as how many hours au pairs worked?; How many children were cared for?; Were pets also looked after by the au pair?; How much pocket money was thought appropriate?; Did employers use an au pair agency?, word of mouth?, websites?; Did they enjoy having an au pair in the house? and so on.

Making contact with employers for interview

We were surprised to find that making contact with employers of au pairs who were prepared to be interviewed about their experiences was one of the most challenging aspects of the project. The surprise was because based on experiences of previous research projects we had assumed that we would be able to make contact with employers more easily through agencies, personal contacts, advertising on websites such as mumsnet and through snowballing. However, we found this was not the case and it was a frustrating aspect of the project. We offered a £10 M&S voucher to all those interviewed.

We contacted au pair agencies to ask if we could send out a flier to employers on their books but we had no response.

We responded to ads placed by prospective employers explaining the purpose of the study and asking if the advertiser was interested in participating. However, the RA on the project received a number of rude, abusive and rather threatening responses to this tactic as people resented the ‘intrusion’ from an outside party. This tactic was abandoned.

Social media/facebook
We attempted to make contact with employers by posting ads on facebook pages used by families seeking an au pair and by placing ads on sites such as mumsnet and netmums. We had no success with the facebook postings and very limited success with the ad on mumsnet.

Personal contacts
We found the only reliable method for recruiting employers willing to be interviewed was through personal contacts and word of mouth. This was limited to the number of people who fit the criteria for interview that we knew or could make contact with through friends and contacts. Also, for one of the two interviews who drew on contacts through her neighbourhood network and through her children’s school, this risked blurring boundaries between personal and professional and meant that she felt uncomfortable with interviewing people she had daily contact with.

Key findings from interviews
The employers were a self selecting group in that is seems unlikely that anyone would voluntarily put themselves forward to be interviewed about hosting an au pair if they felt themselves to be notably bad, unreasonable or exploitative employers. Also as a number of those interviewed were friends or friends of friends of the interviewers it is to be expected that these people might share some of the values and political opinions of the researchers conducting this study. These observations were borne out in that the working hours, wages and conditions and expectations regarding cleaning etc discussed by the employers interviewed were substantially better than those which were typically experienced by the au pairs that were interviewed, and better than the conditions observed in our analysis of advertisements on That is, the employers interviewed appear to be atypical.

The context of hosting an au pairing in the UK
We asked employers where they lived, in what kind of home (i.e. detached house, flat, military base), how many bedrooms they had and whether the size and space in their home had influenced their childcare decisions. We also asked employers whether they lived with a partner or were single, how many children they had, how old these children were and who else lived in their home (eg relatives, lodgers, step children living there part time etc). We asked whether employers worked, and if they did how many hours a week they typically worked and whether their work patterns influenced their childcare decisions. We asked for information about household income and, again, whether this had influenced childcare decisions. We also asked whether decisions to work full or part time were influenced by differing gender roles in the home.

We found:
– Four out of the 15 employers interviewed were single parents. The remainder lived with a partner who was the other parent of the children living in the household. No households had any other members, eg grandparents, lodgers etc.
– Nine of the employers interviewed had two children; four had one child; one had three children and one had four children.
– The youngest child being cared for by an au pair was a newborn baby and the oldest was 12.
– All the people interviewed and their partners but one were employed and the majority of interviewees were employed in professional jobs, except two who were pilates teachers. Of the 15 employers we interviewed, 12 were women and three were men. Two of the men interviewed talked about needing an au pair so that their wife or partner could work. That is, in both cases their very heavy full time work load was not negotiable. Instead an au pair was required to fill in for the mother’s absence. In both cases the mother was responsible for paying and organizing the au pair even though merged household finances made this symbolic more than anything else. The third man interviewed was a single father. He needed the au pair to allow him to work.
– The very typical arrangement we found whereby the male partner worked full time and the female partner worked part time had usually been arrived at after children were born. Before children interviewees explained that they and their partners had been more equal in terms of ambition, career success, pay and working hours.
– It was mostly the case that the female partner arranged and ‘managed’ the au pair so that she (the mother) was able to work part time. The male full-time work role was usually discussed as non-negotiable, even if the male did not necessarily work in an area where he was able to earn more than the female partner.
– It was clearly the case in all people we interviewed that hiring an au pair was a solution to employers’ own employment, rather than an active choice to host a cultural exchange. Many of the people interviewed worked irregular or unusual hours (eg the pilates teachers) and an au pair was a very useful way of ensuring flexible childcare was available. While the majority of people interviewed could be classed as middle class, it was still the case the household income and the size of their homes influenced their childcare decisions. That is, all the people interviewed had hired an au pair because they could not afford a nanny or an hourly paid childcare worker. They had settled upon an au pair because they had the space to host one and they had realized that their relative ‘property wealth’ could be used to subsidise their childcare requirements.
– All people interviewed owned their own home. In all cases the au pair had her own bedroom (although in one case this was across London at an aunt’s house), even if this meant that sometimes children had to share a room.
– In the majority of cases (check exact figure), employers also hired a weekly cleaner. The au pair was commonly required to hoover between cleaner visits, cook for the children and sometimes the whole family, clean the kitchen in the mornings, tidy up children’s toys and do laundry and ironing, rather than being expected to do ‘heavy’ cleaning, as employers put it.

The au pair
We asked how and why the employer had decided to hire an au pair, whether the au pair was their first experience of household employment (eg had they also employed a cleaner, a nanny, a housekeeper etc?). We asked further how they had gone about recruiting their au pair and what they had made of this experience. We asked whether employers had a formal contract with their au pair and if they thought a contract was necessary or helpful.

– As outlined above, all people interviewed had hired an au pair because they needed childcare to allow them to work outside the home and to make their lives easier when they were at home. For example, one interviewee had 18 month old twins (one of whom is profoundly deaf) and a 4 year old. Her husband worked long hours as an architect. She was also trained as an architect but she also had a PhD and was working as an academic. She said the au pair meant her husband did not have to come home for the children’s dinner time but she still had to be there as it was difficult for one person to meet the needs of three infants. She said that the au pair was really freeing up her husband, but also had meant she did not have a ‘nervous breakdown’. Hiring an au pair had been selected over hiring a nanny, a housekeeper or an hourly paid childcarer because it was cheaper and the au pair could also be asked to do household tasks such as laundry, cooking and ironing.
– The majority of employers also employed a cleaner so the au pair was not their first experience of household employment. However, for all but one employer (who had hired a nanny when her children were smaller) the au pair was their first experience of live-in employment.
– People interviewed had found their au pair through word of mouth in their neighbourhood, au pair websites such as au pair world, gumtree or, less commonly, by using an au pair agency. Those who had used an agency usually only did so one or twice before realizing it was easier and more cost effective to skip the agency and make contact with au pairs directly. All those interviewed said that they were inundated with applications if they advertised. This reflects the popularity of London as a destination for au pairs. Employers all explained that they had found it daunting, awkward or difficult when they had first interviewed au pairs but that they got better at interviewing and were more aware of what they were looking for as they became more experienced at hosting au pairs.
– The question of contracts was interesting in that these contracts tend to be symbolic rather than legally significant. This is because au pairs are not ‘employed’ and are not ‘working’ so an employment contract is redundant. In terms of contracts specifying au pair conditions such as hours of work, pocket money etc, there is no one or no government agency charged with specifying what conditions should or should not be and no one charged with upholding contracts. However, some employers had agreed expectations in advance with au au pairs and regarded these agreements as a kind of contract. They did not seem to have much effect on the actual negotiations between host and au pair though.

Hours and pay
Employers paid between nothing (board and lodgings only and £105 a week for between 25 and 30 hours a week on average.
We asked employers questions about why they felt they needed an au pair, what an au pair was for, what the au pair did and how much they were paid. It was clear from the employers’ accounts that they relied upon au pairs to act as low paid childcare and domestic assistants rather than seeing the relationship in terms of it being a cultural exchange scheme. Employers interviewed made clear that they understood the context of the research and understood the potential for exploitation inherent in the conditions surrounding au pairing. Most expressed a desire to be a ‘good’ and thoughtful employer.

Skills and experience
We asked questions about English language proficiency and whether employers saw it as their responsibility to help au pairs improve their English or whether they expected au pairs to already speak English. Employers all expected the au pairs that they hired to be already proficient in English, even though language learning had clearly been an important aspect of the scheme at its inception. They also had high expectations in terms of au pairs being able and interested in preparing healthy and appetizing meals for the children and babies in their care. This was often hard to reconcile with the age and life experience of the young people available as au pairs – ie hosts wanted someone young enough to be comfortable living-in as a junior member of a household, but for this youth to be accompanied by a preternatural maturity, worldliness and ability to cook, look after a house and know how to care for children.

Legal framework
There was very little awareness among employers of the legal framework – or absence of one – around au pairing. Some employers had a vague sense that they should not host au pairs from beyond the EU, although they were not always clear why. There was also a vague sense that it was not a good idea to host an au pair for longer than one year at a time, but again the reasons for this were not clearly explained to interviewers. Employers tended to gain a sense of the ‘right’ way to treat an au pair from reading, often outdated, information posted by au pair agencies and/or by online forums such as mumsnet and/or through talking to friends. This meant that the information was not always accurate. The situation is not helped by the fact that there is no legal baseline or legal position in relation to the employment of au pairs.

Childcare philosophies
In terms of employers’ account of their childrearing and childcare philosophies, the interviews tended to bear out the findings discussed in earlier literature on childcare cultures in the UK. That is, that childcare is not seen as the responsibility of the state and nor are extended family networks commonly relied upon. Rather, nurseries for babies and young children are seen as less suitable then one on one care in the child’s own home. This discourse is backed up by the considerable expense of nurseries in the UK, especially when compared with the cost of hiring an au pair. Some employers had used a nanny when their children were babies and infants and switched to an au pair when their children were older but for the majority of employers interviewed au pair seemed to be exchangeable for a nanny and was seen as preferable to nursery or after school clubs etc.

Part of the family?
Employers interviewed all did understand that the au pair was supposed to be treated as ‘one of the family’; like a ‘big sister for the children’, as one put it. After having acknowledged this, employers did also recognize the potential for exploitation in the au pair arrangement. However, it was also clear that the employers did not generally experience having an au pair as having gained a family member. Relationships were often strained and awkward (around the use of the kitchen or items of food in the fridge, for example). Some employers discussed experiencing hosting an au pair as being akin to suddenly having to look after an extra child. Others felt they were being used as convenient and free lodgings from which au pairs conducted their studies and social lives, with a bit of grudging child care thrown in. Employers talked about being ‘friendly’ to their au pairs but this seemed distinct from actually being friends with the young person in question. That is, employers were happy to be superficially friendly to the au pair but if the employer had their own friends around for dinner, for example, there was a sense that the au pair should know their place and disappear. The majority of those interviewed did explain that they understood that the au pair was supposed to be on a cultural exchange and a number did attempt to facilitate this by arranging trips, helping to enrol the au pair in language courses and other studies etc. Underlying this there was, though, still the sense that the fundamental transaction taking place was one of employment – i.e. the employer wanted someone to perform childcare and housework and the au pair wanted a position in the UK. This exchanged seemed more fundamental than the ‘one of the family’ or cultural exchange window dressing. That said, a number of employers had remained in contact with former au pairs and it was clear the genuine bonds of friendship had been formed.

Meal times and eating
Food and eating emerged as important issues for employers as well as au pairs. There were very different practices amongst interviewees in terms of who the au pair ate with – whole family, just children, alone. Au pairs eating food that hadn’t been intended for them could be a flash point, it did not take much for the employers to feel that an au pair had over-stepped the mark in terms of use of the kitchen or availing themselves of food in the fridge and this could then shape attitudes towards au pairs and relationships between hosts and au pairs. Those hosts who had had a number of au pairs described negotiating around food and eating to be one of the things they had learnt with experience.

Our interviews were largely with hosts who seem to be atypical in terms of the conditions and experience they provide to au pairs. They were concerned to be reasonable and most actively engaged to at least some extent in providing some element of cultural exchange to au pairs. However, even in these conditions relationships between host and au pair could be difficult to negotiate and small issues, such as eating the ‘wrong’ food, could become flash points. While hosts were dependent upon au pairs to allow themselves and/or their partners to work and were generally appreciative of their au pairs they still did not explicitly understand what their au pairs did as ‘work’.

Findings from interviews with au pairs

Aims of the interviews

We aimed to interview 40 au pairs. We initially expected to interview 20 au pairs in London and 20 au pairs in one other region of Southern England – eg around Cambridge. However, we soon found that rather than attempt to locate interview subjects by geographical area it was more practical to make contact using internet sites and then interview the chosen subject using Skype. This meant that we had a greater geographic spread than we had originally anticipated having. We did not limit ourselves to au pairs of any specific nationality or age group. Rather, we wanted to find out more about what the different countries au pairs in the UK had come from, how they had traveled to the UK and why they had decided to become au pairs at all and au pairs in the UK more specifically. We asked how old they were and whether they had previously lived at home with parents or other family members. We were interested in how much language acquisition was an important factor in deciding to become an au pair and whether au pairs perceived what they were doing as a life-stage event or a migration. That is, were they intending to be an au pair in the UK for six months to a year and then return home or move somewhere else or were they using the opportunity to be an au pair as a chance to migrate to the UK? We wanted to know the level of formal education the au pairs we interviewed had achieved and whether they had any particular interest in or training in the care of children. We asked au pairs how they had found out about being an au pair in the UK and how they had organized their placement. We also wanted to learn as much as we could about the experiences au pairs had had or were having. How many children were they asked to look after? Were they also required to look after pets or do other duties? How much housework was involved? How much pocket money were they given and was it enough to live on? Did they take on other jobs as well? What were their employers like? What was their accommodation like? What did they think of the UK? Did they feel exploited or like a worker or a member of the family? How had the experience of au pairing in the UK shaped au pairs’ own parenting philosophies and ideas on motherhood, fatherhood and family life? What did they think of British food and dining habits? All these questions and more helped to give us a picture of what it is like to be an au pair in the UK.

Making contact with au pairs for interview

It is important to note that the term ‘au pair’ is officially meaningless in the UK context so in seeking to interview ‘au pairs’ we relied upon finding people who had advertised their services under the ‘au pairs and nannies’ category on or who were otherwise identified as au pairs by the categories used by websites and agencies, people who self identified as au pairs or were identified as au pairs by employers. We also relied upon Facebook and other social media groups where people self identified as au pairs to make contact with individuals.

We offered a £10 M&S voucher to all those interviewed.

All interviews were recorded, transcribed and then coded using NVivo.

Some interviews were conducted face to face (in the interviewer’s home or in a café). However, we were making contact with people using the internet and the people we ‘met’ were often located a long way from central London. Also, au pairs have real limits on their free time. We found it was easier for the au pairs to offer to interview using Skype. Therefore the majority of the interviews were conducted using Skype.

We attempted to make contact with au pairs by contacting au pair agencies and explaining the aims of our study but agencies contacted by email did not reply. We decided it would be more productive to attempt to contact au pairs directly and so used gumtree as well as facebook groups to ‘approach’ au pairs over the internet.

Gumtree has a section where people looking for work can advertise their services to prospective employers. We aimed to make contact with people who advertised themselves as au pairs looking for a first or a new position by responding to these ads and explaining the purpose of our project. In fact, the majority of au pairs who advertised themselves on gumtree were already working as au pairs and were seeking a new position. A number of au pairs contacted in this way were disgruntled with their current au pair position and that is why they were actively seeking a new position. This should be noted in terms of the effect how we found the au pairs interviewed had on the outcome of the interviews.

Social media/Facebook
Au pairs and prospective au pairs have an active online presence. There are numerous talkboards, Facebook groups and individual Facebook pages through which people look for au pairs or host families. Au pairs also use these pages to make contact with each other in order to discuss their experiences and to make contact with new friends. The pages are used to organize social events for new arrivals in cities and towns all over the world and to ask advice of each other. We used Facebook to post messages to groups and to individual au pairs asking people to contact us if they were interested in being interviewed.

Personal contacts
Some of the au pairs interviewed were working for friends or community contacts of the interviewer. This was not ideal, though, as the au pairs in question tended to be part of the broader school community of which the interviewer and her children are part. Interviewing these au pairs blurred personal and professional life in a way that was sometimes awkward.

Key findings from interviews

The interviews with au pairs were very successful in that all people interviewed engaged wholeheartedly with the interview process. The people interviewed understood the questions and most people interviewed had a very good idea why we were engaging in the study and were interested in the background to the research. Language barriers were not generally a problem as there is a strong preference for people who already speak English among employers of au pairs in the UK.

Age and nationality
The au pairs interviewed came from 15 different countries, all in Europe. The most important countries numerically were The Czech Republic (6 interviewees), Germany (6 interviewees), Romania (6 interviewees) and Spain (5 interviewees). Between them these four countries were home to more than half (23 /40) our au pair sample. A total of 18 of the au pairs interviewed came from Western Europe (Germany, Spain, Italy, Finland, France and Denmark – in order of importance) and 22 came from Central and Eastern Europe (Romania, Czech Republic, Bulgaria, Poland, Estonia, Slovakia, Slovenia and Latvia).

All the au pairs interviewed were aged 18-29 years. Fifteen were 18-21 years old, thirteen 22-25 and ten 26-30 (n=38). The au pairs from Western Europe tended to be younger than those from the East, for example all the au pairs from Germany were 22 years old or younger, whereas only one from Romania was under 25. See t table above for a detailed breakdown of age and nationality of au pairs interviewed.


Hours worked etc
The interviews assisted in quantifying the experience of au pairs interviewed in terms of the amount of time they had spent as au pairs in the UK and elsewhere; how many hours a week they were required to work; how much they were paid. In this the interview data reinforced findings from the analysis of ads for au pairs placed on The strength of the interview data is that we were able to ask au pairs to elaborate on the number of hours worked by asking what exactly they were required to do (i.e. were they cleaning and if so what kind of cleaning? If they were doing childcare, what was expected of them? How were the treated by the family? Were they happy?). We found that au pairs were commonly relied upon to act either as fulltime child carers for pre-school children and/or to fill the gaps at the beginning and the ends of each day so that parents could fulfill employment requirements. Au pairs also commonly cleaned the kitchen after meals, did laundry, hoovered the whole house at least once a week, shopped and cooked for children (and sometimes for the whole family), walked and groomed pets, ferried children to swimming lessons, ballet, kumon etc.

As the table above shows, among the 33 au pairs who gave us information about their pay there was a wide variation, from no pay at all to £200 per week. Eleven au pairs were given up to £75 per week in pocket money; fifteen received £76-100 per week; six £101-150 and one over £150 per week. The hours worked also varied from just three per day to over 60 per week and those who were working the longest hours were by no means the highest paid. In fact, as the table shows there does not seem to be any relationship between the hours that were worked and the pocket money received.

Worker or cultural exchange?

The table above provides a summary of nationalities, formal education levels reached, length of time spent as an au pair, hours worked, pocket money paid, whether English lessons were being done, whether the au pair intended to stay in the UK or go home and how the au pair was contacted. The overall impression gained by the interviewers was that au pairs felt exploited and vulnerable in the UK. Even those who claimed to be ‘happy’ and to have enjoyed or to be enjoying their experiences as an au pair spoke of very long hours, unpredictable and self -centred employers, difficult and sometimes aggressive children and isolation.

The au pairs who claimed to be happy were in the minority. Far more common were tales of exploitation and sometimes verbal and physical abuse. The vast majority of au pairs interviewed explained that they felt they were being made to do childcare and house work on the cheap. This was either because the employers were themselves cash strapped and stringing together a household/paid work survival strategy or because employers were mean and had realized calling a household employee an ‘au pair’ allowed for the extraction of much work for little money. Either way, the people employed as au pairs had no way to fight back against their perception that they were actually working rather than being hosted on a cultural exchange.

Au pairs were also intimately involved in the lives of their employers and those of their children and in some cases this intimacy was experienced as a kind of abuse by au pairs. This ranged form exposure to employers’ alcohol problems to having to step into the breach when employers left au pairs alone with children for several days at a time while the employer left the country (or was arrested after a night out, in one case). There were also examples of au pairs feeling as though parents were projecting their frustrations on to them and incidences of cultural gaps in terms of how children should be fed, disciplined, included in family life etc.

Parenting philosophies
Au pairs made judgments about the parenting philosophies and parenting skills of their employers and this affected how they felt about their own families and how they imagined their own future families. Many au pairs found the way in which UK families often feed children separately from the adult dinner time very alienating, and where there was a requirement that the au pair eat with the children this was sometimes experienced as infantilizing. Some au pairs found their charges cute and charming but this was by no means the consensus as there were many stories of violence and lack of respect.

Possibilities for redress by au pairs
Au pairs interviewed had no sense that there were any possibilities to improve their conditions through collective action or through outside agencies. The only thing they could do was to leave and either return home or find another household to employ them. For some of those interviewed this was relatively straightforward, but for others the period between telling the employer they wanted to leave and moving to a new household had to be very carefully negotiated if the au pair was not to face a period of homelessness. Also, the older, more experience and savvy au pairs were better equipped to confront exploitative employers than were the very young and inexperienced young people.

Life stage experience or migration?
There seemed to be a real difference identifiable between those au pairs who were treating the experience as a time out from their lives – ie as a break between school and university or between school and work etc ¬– and those who had a more open-ended view of the space au pairing was filling in their lives. There was a correlation between being from Western Europe and being more likely to be planning to return home as well as higher levels of formal education and being more likely to be planning to return home. Also, some of those from wealthier backgrounds were being partially supported financially by their parents and this meant they were less exposed to exploitative conditions than those with fewer options. However, even those for whom au pairing in the UK represented a ‘break’ reported feeling like they were working and feeling as though they were over worked and under paid.

Au pairing after the au pair scheme…Research finds widespread exploitation of au pairs and a system open to abuse

Our two-year ESRC-funded research project collected data from au pairs and host families and the findings are published today (16 October 2014). We found that the average au pair in the UK works over 38 hours a week, although some are expected to work for up to 70 hours, with expected duties sometimes including caring for elderly relatives, or helping out in family businesses. Average pay is £108 per week, but 14% of au pairs do not receive the £85 a week recommended by the British Au Pairs Agencies Association.

Au pairing was traditionally supposed to offer young people the opportunity for adventure and cultural exchange, but most hosts interviewed conceded that meeting their childcare needs was their motivation for employing an au pair and many au pairs felt that their hosts were not interested in providing opportunities for cultural exchange. 44% of those advertising for au pairs expected prior experience, and 26% were only considering applicants who are already in the UK, showing that the increasing reliance on au pairs is leading to a decreasing differentiation between au pair and nanny roles. Many au pairs are significantly older than the typical image of someone in their late teens or early 20s, with the economic situation in southern Europe spurring those in their mid-late 20s on to improve their English and ‘wait out the crisis’ or use au pairing as a first step to more permanent migration.

The key findings from the project are officially ‘launched’ on the Birkbeck website today.There is also a pdf document that can be downloaded from here.

  • The research was carried out by Dr Rosie Cox and Dr Nicky Busch