Gender, class and space in the field of parenthood: comparing middle-class fractions in Amsterdam and London

This paper argues that becoming a parent/carer can be seen as a new field of social relations and suggests how gender is the key mechanism in the reconfiguration of class relations in this field. By conceptualising parenthood as a field, that is a social world with specific stakes and rules, this study suggests that residential decisions and strategies developed by different middle-class households do not solely depend on their class habitus, but also on gendered positions and dispositions in respect to division of labour, child care and school choice. Drawing on interview data from London and Amsterdam, this study re-addresses the issue of middle – class time-space trajectories at a specific period in the life course. We contend that the middle classes are not just differentiated by various orientations of capital (economic versus cultural) but that interaction of class and gender is also key for understanding practices of the middle classes as they enter the field of parenthood. These practices are strongly influenced by labour market and welfare regimes (as the Netherlands/England comparison makes clear). In the new field of parenthood the work of realigning class habitus (through social reproduction) is highly gendered, but to different degrees that are made evident in the different neighbourhood settings. In terms of urban space this points to the significance of the particular neighbourhood structure and opportunities of the city as a whole as well as a more active idea of the role of space in the particular working out of class and gender in specific neighbourhood contexts. Urban space is a situating framework and an active process in trajectories of social reproduction.

< h t t p:// dx. doi.o r g/ 1 0. 1 1 1 1/ t r a n. 1 2 0 7 3 > Pl e a s e n o t e: C h a n g e s m a d e a s a r e s ul t of p u blis hi n g p r o c e s s e s s u c h a s c o py-e di ti n g, fo r m a t ti n g a n d Thi s v e r sio n is b ei n g m a d e a v ail a bl e in a c c o r d a n c e wit h p u blis h e r p olici e s. S e e h t t p://o r c a . cf. a c. u k/ p olici e s. h t ml fo r u s a g e p olici e s. Co py ri g h t a n d m o r al ri g h t s fo r p u blic a tio n s m a d e a v ail a bl e in ORCA a r e r e t ai n e d by t h e c o py ri g h t h ol d e r s .

Introduction
The transition to parenthood is a key phase of the life course and in the reproduction of class relations. It is a particular pinch point that raises the stakes in class reproduction through the adjustments that parents have to make in the labour market and through concerns over social reproduction (and especially schooling). Parenthood can be conceptualised as a 'field' in Bourdieu's terms (Boterman 2012): it is a social world with specific stakes and rules. The elements that potentially unsettle the class dimensions of a habitus-field balance also raise the stakes in gender relations. This can be seen in the case in middle-class households in which both partners have professional careers prior to becoming parents. It is also middleclass parents that have been identified as being the most active in developing schooling strategies for their children's education (Ball 2003;Butler and Hamnett 2007;Reay et al. 2011). These adjustments over employment and schooling take on a spatial aspect in that it is middle-class professional households that have the greatest capacity to try to accommodate their changing needs through residential moves and neighbourhood selection. Drawing on 176 in-depth interviews from two studies of London and Amsterdam, this paper will re-address the relationship between class, gender and urban space. It will suggest that the residential decisions and strategies developed by different middle-class households do not solely depend on their class habitus, but also on gendered dispositions in respect to division of labour, child care and school choice. These are influenced in turn by the sociopolitical and national institutional arrangements that partially structure labour markets, education and childcare. The Netherlands and the UK present contrasting institutional contexts that might suggest different gender outcomes in middle-class households (with differential impacts on the social geography of the city as a result). This comparative analysis of the middle classes in Amsterdam and London also suggests how in the field of parenthood gender is constitutive of forms of class reproduction in urban space.

Gender, class and space
One aspect of the literature on the relationship between gender and space has been concerned with analysing the impact on gender relations of different national social policy regimes (Duncan and Pfau-Effinger 2000;Sainsbury 1994Sainsbury 1996. Pfau-Effinger has suggested how varying gender cultures can be discerned in different European countries that underpin and explain these social policy differences (Pfau-Effinger 1998). Analysing the sub-national scale, Duncan and Smith (2002) suggest there are inter-and intraregional contrasts in gender cultures within the same welfare regime (Britain). Further analysis of the relationship between gender and class in the rationalities of motherhood (Duncan 2005) shows inter-and intra-class distinctions in the cultures of motherhood and the women's relationship to paid work during the parenting/caring phase of the life course. In the latter study Duncan points to distinctions within the middle classes between primarily motherhood-favouring suburbanites and primarily worker-favouring gentrifying partners. Those middle-class households that make the transition to parenthood and choose to stay in the city (rather than suburbanise as traditional urban models would suggest) represent a particularly interesting test case of the interrelationships between gender, class and space. Some of these elements have been dealt with in an earlier literature on gender and gentrification to which we now turn.

Gender and gentrification
The inter-relations of class and gender in gentrification have been considered from a number of theoretical and empirical perspectives. As gentrification was gathering pace in the 1980s, Beauregard (1986) pointed to the significance of delayed fertility of professionals that put extended emphasis on the social opportunities offered by certain neighbourhoods in the central city in the search for friends and potential future partners. Changing macro divisions of labour and increasing female participation in the professional labour market were the driving force of gentrification according to Warde (1991). Consideration of women's careers offered a plausible explanation of the two prominent forms of gentrificationcommercial, capital-intensive new-build developments, and the sweat-equity gentrification of more historic working-class neighbourhoods (Clay 1979). The increase in professional women delaying fertility or choosing not to have children explained a large part of the demand for new-build apartments in the central city. The other manifestation of gentrification involving sweat equity, Warde argued, could be seen as an outcome of the competing pressures on professional women in (lone or dual earner) family households. Patriarchal power still bore down on this situation in that cutting down on commuting time was still more significant for mothers than fathers, labour market areas for women tended to be smaller and there were persistent gender divisions in the organisation of formal and informal childcare. The management of the competing tasks and constraints has to be seen in the context of increased labour market participation for women overall and the way that 'women organise their trajectories through paid and unpaid labour' (Warde 1991, 229). Warde pointed to lifecycle effects that potentially link these two modes of gentrification through the changing constraints on women's careers and this is of particular relevance for our current discussion. He concluded that 'gentrification is a process of displacement of one class by another, but its dynamics are better understood as originating in changes in the labour market position of women ' (1991, 231).
Rose (1984 1989) saw gentrification also comprising dual-earner households for whom the central city permitted a more equitable division of formal and domestic labour. She emphasised the importance of the proximity and mix of services and facilities in the central city, from the perspective of support to single professional women (with or without children) in parttime or unstable employment. For these 'marginal gentrifiers', the central city was an affordable resource to manage the multiple tasks in the daily round. Bondi (1991) looked at the changing role of women in economic and demographic terms and the connections to gentrification (and its divisive impacts on working-class women displaced or marginalised in gentrifying neighbourhoods). She argued that gentrification is a process of gender (as well as class) constitution. Bondi suggested how any such research should be sensitive to the historic and particular geographical contexts in which this process of gender constitution (in relation to class and ethnicity) takes place. In a later paper (Bondi 1999) she points to the significance of life-course factors in assessing the relationship between class and gender in gentrification, along with the need to look at these relations in different neighbourhoods and, crucially, in different cities. The majority of her respondents in Edinburgh expressed egalitarian views about gender roles and employment, but gentrifiers with children suggested that the pressures and practicalities of raising children severely challenged these aspirations.
In our argument about parenthood as a field, these disruptive and challenging conditions are seen as more widely significant. In a cautionary note commenting on Warde and Bondi's contributions, Butler and Hamnett (1994) argue that these more egalitarian gender practices are limited to a distinct (and relatively privileged) fraction of the middle classes: the new urban professional middle class. The empirical evidence from the London and Amsterdam studies includes the middle classes living in the suburbs and exurbs as well as in the central city. They are all part of the upper professional/ managerial fraction of the middle class but show some variation in lifestyle and political and social outlooks across this fraction. Child rearing is an emphatic period in the ongoing construction of gender relations within the household. The context of the particular city in which the negotiation of these time-space constraints occurs is critical. A number of Dutch studies demonstrate how 'family gentrifiers' are now an established group in Dutch urban landscape (Boterman et al. 2010;Karsten 2003Karsten 2007Van den Berg 2013). Karsten shows evidence of more equitable divisions of labour for some households (with both partners working four days a week) but also that more generally inequalities in the gender balance between formal and informal work (including parenting) persist. Karsten's work highlights the connections and interactions between residential neighbourhood, class and gender constitution, but she deals only with those families that stay in the city, those who remain as gentrifiers. However, the residential mobility literature points to the significance of fertility and family formation in prompting changes in residential trajectories, the classic trajectory being one of suburbanisation (Bell 1968). As this literature has shown, residential choice is strongly interconnected with the life course, and specifically with family formation and parenthood (Clark and Dieleman 1996;Rossi 1955). To what extent do gentrifiers stay in the city or leave the city when they have children? The transition to parenthood can unsettle habitual patterns of class reproduction. Previous research has investigated some of the continuities and discontinuities in neighbourhood trajectories of gentrifiers in terms of what they say about different fractions of the middle classes (Boterman 2012;Bridge 2006) and we develop this approach in the current paper.
Unsettling the habitus: parenthood, gender, class and space Savage and colleagues have argued that 'people are comfortable when there is a correspondence between habitus and field, but otherwise people feel ill at ease and seek to movesocially and spatially ' (2005, 9). Most Bourdieu-inspired scholars tend to agree that habitus essentially generates spatial divisions between various groups (Atkinson 2006;Butler and Robson 2003b;Savage 2010;Watt 2009). As Butler and Robson (2003a) have argued, these spatial divisions are based on practices in various fields such as education, employment, consumption and housing that are associated with various orientations of capital. Residential choice is therefore considered to be the outcome of the process of aligning habitus (that is the relative amounts of social, cultural and economic capital) and field. In this strain of literature relatively little attention is given to residential relocation associated with changes in the life course and the effects these have on the interrelationship between gender and class. Although many UK scholars of the urban middle classes pay attention to the relationship between school choice and residential choice (Ball 2003;Butler and Hamnett 2007;Reay et al. 2011), these are very much concerned with pointing out differences between middle-and workingclass parents and between fractions within the middle classes. In a qualitative longitudinal study of gentrifiers in Bristol, England, Bridge problematised this. He wondered, when middle-class households move out of a gentrified neighbourhood (usually in the hunt for 'better' schools), do they simply cease to be gentrifiers and, if so, what does this say about claims that gentrification is the cultural display of a distinct new middle class? (2003,2546) He concludes that the way in which the habitus is articulated in a field is dynamic throughout time and space and is also highly geographically contingent. The middle classes have various time-space trajectories, which are informed by their habitus and by the local configurations of the fields (Bridge 2001). Yet, although Bridge has introduced a longitudinal perspective and emphasises that we should think in terms of trajectories, his work primarily considered habitus in terms of various forms of capital and focused mainly on the relationship between school choice and residential choice.
Drawing on longitudinal data Boterman (2012) demonstrates that urban middle classes in Amsterdam have changing housing preferences and practices in different stages of their lives. This study confirms much of Bridge's conclusions that residential practices involve trade-offs between various forms of capital (including time) that are informed by their habitus. Although Boterman's study also primarily stresses the class dimensions (forms of capital) of habitus, his work also suggests that entering the field of parenthood can unsettle habitual patterns of class reproduction. Furthermore, he points to the consequences for the positioning of men and women in various fields, most clearly in the field of employment. For example, different gender ideologies result in various divisions of paid and unpaid work. These different divisions have consequences for the financial and time-space budget of families and thus influence residential practices, but also for the position of men and women in other fields. Decisions concerning child care and school choice seemed to be made primarily by the mother who is primarily responsible for caring tasks. This resonates with the work on gender and gentrification cited above which emphasised the role of the labour market and how this bears down differently on men and women. In spite of the evidence that time-space trajectories of middle classes may not just be about class but also about gender, few studies of middle-class habitus have systematically treated gender at an equal level with class. However, the mechanisms by which any possible disruption to the habitus is managed and stabilised is as much to do with ongoing gender relations (and their spatial impacts on neighbourhood trajectories) as it is with class per se.
This paper aims to revitalise some of the discussions on gender and gentrification by developing Bourdieu's (1984) concepts of habitus, field and capital. The paper considers parenthood as a field in which gender relations are the critical mechanism in the realignment of class habitus and field. We explore how gender impacts on time-space trajectories of the middle classes and how these trajectories simultaneously inform and reproduce gender roles. Furthermore, by looking at the way in which the use of urban space, notably the neighbourhood, also becomes gendered by the transition to parenthood, we also investigate broader linkages between the formations of gender, class and urban space. We now turn to the empirical case of the middle classes in Amsterdam and London to elaborate the argument.

Comparing class and gender in Amsterdam and London
This study compares various fractions of the higher level professional and managerial middle classes in London and Amsterdam. Both cities are characterised by a strongly international economy built on financial and other business services. Extensive parts of the Dutch and British capitals have been affected by processes of gentrification, which have transformed many former working-class areas into middle-class residential and consumption spaces (Butler et al. 2008). Notwithstanding the rise of the middle classes, both cities remain highly diverse in terms of class and ethnicity. The research noted variations in the way in which gender is practised between the UK and Dutch context. It has been argued that gender attitudes and practices are related to 'cultures of care', which include ideas about 'good' parenthood (often motherhood) and about childhood idylls (Van der Wel and Knijn 2006). Obviously these gender attitudes are inseparable from national institutional frameworks and labour and welfare policies. A number of differences between the UK and the Netherlands are of particular importance for this study, especially relating to labour-market policies concerning part-time work and the costs and policy arrangements for childcare.
Due to policies that make part-time work a legal right 1 for all workers in the Netherlands, part-time work is much more common than in other European countries, including the UK. Of all potential workers (age 15-65) about 76 per cent of all women and 25 per cent of all men work part-time. Among mothers this share is even higher; among fathers about the same. In the UK 12 per cent of all men and 43 per cent of all women work part-time (Eurofound 2009). In the Netherlands, the most common division (56%) of labour among families with dependent children is a male full-time, female part-time arrangement (SCP 2012). In the UK exchanging full-time work for parttime work or quitting altogether is even more common among mothers of young children 2 (Sciarra et al. 2004). Participation in the labour market of mothers and divisions of labour are also clearly classed in the two contexts. Higher educated women work more often and more hours. They also tend to leave the labour market less when they have children and return more often (SCP 2012; for the UK see Smeaton 2006). Among the Dutch and British middle classes, divisions of labour tend to be more symmetrical than in the working classes. Yet, among all classes paid work is very strongly gendered, particularly among parents of young dependent children. In the UK mothers work less than in the Netherlands. Perhaps equally importantly, in the Netherlands working only part of the time is not considered a decisive impediment to pursuing a serious career (SCP 2012). Particularly within the public sector, working four days a week still allows employees to reach top-ranked positions. However, in some highly competitive sectors such as the financial services and major law firms, part-time work is less accepted.
The costs and quality of childcare and the national and local provisions that influence affordability have a clear effect on divisions of labour. In the UK context formal childcare costs are relatively high (Viitanen 2005), which has a clearly negative effect on the labour market participation of mothers, particularly of young children. For middle-class families few tax credit provisions exist that mitigate the costs of child care. As Vincent and colleagues (2008) have shown for the London context, middle-class families generally bear about 75-80 per cent of the total costs, while working classes get more significant tax reductions. It is evident that for many families having young children is a serious impairment to pursuing two careers simultaneously.
In Amsterdam the gross costs of child care are comparable with those in London (about €325 per child, per week). The major difference between the Netherlands and the UK contexts are the relatively generous tax provisions that exist for all families with two working partners in the Netherlands. Although tax relief is income-dependent, high incomes (>€200 000) still bear only a maximum of 67 per cent for the first child and get a compensation of 85 per cent for the second child (the situation in 2010). For middle-class households with somewhat lower incomes, the tax credits are more favourable still. As a consequence, for a family in which both spouses work full time, childcare costs are a considerable expense but they do not cancel out the revenues of paid work.

Data
This study draws on 176 interviews with mothers and fathers within various fractions of the middle classes in London and Amsterdam. The dataset is derived from a longitudinal study with middle-class couples in the inner-city of Amsterdam; and a large qualitative comparative study between the middle classes in Paris and London. We draw on 28 interviews, carried out in 2008 in inner-urban neighbourhoods in Amsterdam and 25 interviews in 2010 with the same couples at various urban and suburban locations; and 123 interviews from 2011 with parents or carers (out of 171 resident interviews) in five types of London neighbourhoods: gentrifying, gentrified, suburban, exurban and gated community.
The two research projects were explicitly aimed at uncovering the dynamics of residential choice and social reproduction. In the Amsterdam project the focus was specifically on the impact of having children on the residential trajectories of urban middle classes. The London-Paris comparison selected respondents by their residential location. Although the studies had slightly different research designs, the samples contain very similar households in terms of level of education, income, age and employment, which could all be classified as professional middle class. Because our focus is on the impact of parenthood on residential trajectories, we take the neighbourhood as a starting point of the analysis. We do this for two reasons: first the differentiation between gendered divisions of work in London is very limited, which makes a selection of gender divisions redundant; and second we want to study the wider residential trajectories of urban middle classes and not just offer a static account of gentrification. Throughout the empirical sections we will therefore refer to the residential locations of the respondents when quoting them. We have recoded the transcripts of all interviews using the same nodes in NVivo 9 (London) and Atlas TI (Amsterdam).

Motherhood, employment and childcare
The transition to parenthood has a strong influence on the labour market position of both parents, but particularly mothers. For many of the female respondents in the London neighbourhoods becoming a parent meant giving up formal employment altogether, or for a period time of varying lengths but returning as a part-time employee. For most middle-class women this signified a strong discontinuity with the professionaloften full-timecareers that they had before they became mothers.
In the City ... and I er worked until I went on maternity leave, and had Malcolm and then I wanted to go back part time but they didn't want me to go back part time, so I stopped working and then I had my other son. (Emma, gentrified neighbourhood, Balham, London) In many cases there is no expectation that they will be able to return to work with more flexible or parttime working. In the London neighbourhoods all the discussion was of changes to women's (rather than men's) employment.
I do intend to go back to work. ... I will do reduced hours hopefully; I have to ask my employer whether they're willing to let me do that, because I was doing a full-time role before. (Sarah, gentrified neighbourhood, Balham, London) Whereas in London almost all the discussion was about how women cease formal employment altogether and a few continue working part time, in the Amsterdam sample almost all the women interviewed maintained substantial working careers of 24-32 hours a week. Furthermore, most of the men also reduced their working hours, albeit rarely below 32 hours. The majority of the couples in the Amsterdam sample had a symmetrical part-time division of labour: working four days and childcare at home one day. For most new mothers two things were evident: they wanted to work less (usually one or two days less) because they wanted to spend more time with their children; and they wanted to continue pursuing a serious career. It is often expressed that working four days is just the most 'logical', or just the most 'fair'. From the mother's perspective this 4-4 arrangement is considered the logical outcome of the desire to remain active on the labour market, to have children no more than three days in day care institutions and to share the household and child care tasks equally. Ella: We have three days that we both work. And Dan is together with her on Fridays. So she goes to day care for three days a week. And we both have her one day, a week day, that we take care of her. Or that we are together with her. (Dan and Ella, gentrifying neighbourhood, Bos en Lommer, Amsterdam) [author] Why did you choose a 4-4 division?
Molly: I think it's completely logical that I would work four days. I haven't imagined one single time to work five days; and fewer than four days wasn't an option either. (Molly and  Roberto, gentrifying neighbourhood, Westerpark, Amsterdam) A pervasive theme from the London interviewees, regardless of type of neighbourhood, is the impact of the cost of childcare on women's careers. The association of women with childcare is registered in the particular ways that the term 'mum' or 'mums' is deployed and the absence of this kind of framing, or its equivalent, in Amsterdam.
A lot of the mums actually don't work ... with the cost of child care they tend to give up work and then start working again perhaps when the children go to primary schools. But it's probably mixed but it's half are not working and half are working but most people I would say 100% work part-time, I'm planning to come back myself just two days because it just gets so busy and it gets very expensive as well when you have five days of child care with three children. (Emily, gentrifying neighbourhood, Peckham, London) In Amsterdam the decision to work is also influenced by the ambiguity over the perceived 'quality' of caregivers. Most people see the child care facilities as an indispensable, but not ideal solution. The specific 4-4 division of labour is among things an expression of the half-heartedness of middle-class parents towards these institutions.
John: Well, we both wanted her no more than two days in daycare. That was just a feeling. You choose this deliberately and we thought three days, well that's a bit too much. (Daisey and John, gentrified neighbourhood, Oostelijk Havengebied, Amsterdam)

Fatherhood, employment and childcare
Whereas discussions of parenting in relation to employment with female parents were charged with practical management issues and emotional dilemmas, those of male parents (or accounts of partners from female parents) displayed an overall absence of these issues. Particularly, across the London neighbourhoods there is an absolute norm of male full-time employment. This encompasses a range of public-and privatesector occupations. It also includes a range of types of middle-class neighbourhood, from the more liberal or progressive households with more egalitarian perspectives on childcare, through to the more conservative middle classes where gender divisions over work and views of parenting conform to a more traditional norm.
So, so we live two completely separate lives you know, in the week really. So, me and the children and him sort of working away so... (Charlotte, gentrified neighbourhood, Balham, London) No, he works in finance ... but it is long hours so ... so he doesn't get to see Toby very much in the week which is a shame. (Fiona, gentrified neighbourhood, Balham, London) Here the norm of long hours employment for the father means that not seeing his son on weekdays is a matter of regret, rather than being an acknowledged inequality in childcare responsibilities, or being an issue invested with moral overtones as is frequently the case in relation to 'being a mum'.
I mean I will tend to leave at about quarter past seven and my wife's got up and the kids are just getting up, so they, they, normally now ... My wife is she's a full-time Mum. (David, suburb, Berrylands, London) There are cases where the norm is criticised but is still seen as a fait accompli. It can also affect the female partner's life beyond childcare hours, as this next quote attests.
Too many of the husbands work stupid hours, and seriously some of them are 10 o'clock commuters at night, an awful lot of them don't get home in timefor their wives to, go out to things like book club and things so I find it a little bit, it's quite odd. (Tina, suburb, Berrylands London) For many middle-class fathers in Amsterdam it is generally a matter of course to assume a serious role in the household and take care of the children. The general attitude for most men is to share the paid and unpaid responsibilities equally. Many fathers start working part time or squeeze a full-time job into four days at the office and work the rest at home. Roberto: 4 days each. That's just the most fair. Yeah, it's just split in the middle, I think.
[author]And you wanted to do it fair? In Amsterdam both parents are thus committed to caring tasks as well as paid work. As the quote suggests, however, expectations about work and working hours are not the same for mothers and fathers. Some employers are more traditional than others and differentiate between men and women. This also implies that educational and employment trajectories already prefigure some of the constraints and opportunities when entering the field of parenthood.

Parenthood and urban space
The elements that interconnect gender, employment and parenting have an impact on urban space, which in turn helps consolidate or reconstitute the relations between class and gender. One element is residential choicefrom the scale of space within the home, to the nature of the neighbourhood and the overall location in the city. In Amsterdam most families are dual earners and residential location is often a compromise of the relative location of the workplace of both spouses. The time-squeezed young family households are clearly strategising in order to minimise commuting times but in ways that are more equitable in gender terms than in the London case: Natalie: So ideally it means that he has to find a job in Amsterdam.
Robert: Yeah exactly. Natalie: ...because, since I work in The Hague we cannot both commute.

Robert: Precisely.
Natalie: If it doesn't work out to find a job here we might even consider moving. You can only have one commuter in the household, not two. (Natalie and Robert, gentrifying neighbourhood, De Baarsjes, Amsterdam) It is often described as a luxury to be able to bike to work and only spend 20 minutes or so on commuting each way. In addition to length of commuting times, living close to work also provides a clear sense of control over time, which is particularly important in respect to opening hours of child care facilities, for instance (Schwanen 2006). This control over time and managing the time-space budget of the family is a key point for understanding why many families choose a central location. In the inner London neighbourhoods it is the commuting time of the male employee that dominates.
A second important gender dimension of residential location is the impact on women of residential moves that are prompted by parenthood.
Molly: Well, if you have a child it's already pretty hard to be spontaneous. If you live far out and you want to have a cup of coffee I'll have to bike for 30 minutes just to come to the train station, so not even there, so if I want to meet with a friend I'll have to travel quite far. Then you just don't go. So it's hard because you are a mother and because of the place where you live, where you do not feel connected to at all. I would really be very lonely there. It seems really terrible to me. (Molly and Roberto, gentrifying neighbourhood, Westerpark, Amsterdam) Molly is afraid that moving into the suburbs would cut her off from the social life and working life and that she will be lonely. For her being in the urban buzz provides her with a clear sense of belonging. As she and many other parents describe it, the meaning of the residential environment becomes much more important when you have children. 'Classic' studies on the residential trajectories of middle classes emphasise that characteristics of the dwelling itself tend to be prioritised, but many of the families in the Amsterdam context report juggling with a range of responsibilities, which lead to a broader variety of residential trajectories. These sentiments are also expressed in London (where feelings of alienation were more evident than in Amsterdam): I didn't really want to live in ...um ...suburbia, I wanted to live somewhere where I felt like I was part of the city ... your horizons reduce when you have lots of very small children ... living in central London meant you were actually part of the community and things were happening in the day and if you walked out the door you'd come across people that were working in offices, people that were working in factories, people that were, school children mums and children, retired people and some from all classes of society and that was something I really relished and enjoyed. (Sarah, gentrifying neighbourhood, Peckham, London) For some, an inner urban location is preferable to counter the possible feelings of isolation that come with being a mother of a young child, that certain respondents felt may be emphasised in a suburban or exurban location. The London study indicates some of this in several respondents' accounts from those living in the city as well as in the suburban or exurban locations themselves, or through the experience of moving to suburban or exurban locations.
So that was very difficult to manage, and also I was pregnant with my first son, and so when I came down I didn't particularly know anybody, I had three young children, very isolated and lonely. (Sasha, commuter village West Horsley, London) I suppose there is a trade-off because it can be a bit of a soulless place as well though. It can be pretty dead feeling, as much as it being safe it can feel ...Nothing ever happens. (Joy, suburb, Berrylands, London) Elsewhere the suburb or the commuter village is valorised as the ideal location to bring up children. In the London suburb this was overwhelmingly to do with access to good schools as well as the amenity of a large garden that comes with semi-detached 1930s suburban dwellings.
Becoming a parent means becoming invested with responsibilities for social reproduction, especially through the pre-school activities and environments and choice of the child's school. Issues of school choice, especially of middle-class choice and strategies in education, have been the subject of considerable research (Ball 2003;Boterman 2013;Raveaud and Van Zanten 2007;Reay et al. 2011). As well as the arguments about the class-based nature of school choice there are important gender effects from the management of social reproduction via education. The evidence suggests that it is women that are most active in school selection (Reay 2000). Women are the main managers of the children's experience with the school (the school run and liaison with teachers and support to academic activity out of schoolespecially homework). Women are also the key members of the 'hot grapevine' of knowledge (Ball and Vincent 1998) about schools and within schools. This is a largely UK-based literature and to some extent contrasts with the accounts from Amsterdam where the mechanics of schooling appear to be shared more equally between partners. In terms of neighbourhood choice, there is a significant difference between the English school choice system, which is premised on choice but where postcode does matter in terms of access to good schools, and the system in the Netherlands, where neighbourhood location has little or no bearing. Whereas in London education is key to understanding residential trajectories of parents/ carers, most families in Amsterdam see education only as one factor out of many. To some extent this relates to the particularities of the Amsterdam schooling landscape that emphasise free school choice and a relatively weak tie between neighbourhood and school. The trade-off of location versus schooling, which is very common in London, seems to play only a small role in Amsterdam.
As a respondent in one of the London commuter villages expresses it: It's an interesting dynamic because when you're involved with the school, which you have to be when you've got children, everything is done by mums, and that's what you'd expect, but almost to the exclusion of dads altogether ...

(Clara, commuter village West Horsley, London)
My husband is beginning to talk about moving to the country ... Er ... I don't really want to move, er but, the reason there's lots of private preps [preparatory private schools that link into well-known public schools] round here. (Rose, gentrified neighbourhood, Balham, London) Overall, in terms of one of the key mechanisms of class reproduction (schooling and all that comes with it), there is a strongly gendered dimension in London and one in which neighbourhood selection is critical, which contrasts with a more even spread of this responsibility between partners in Amsterdam and where schooling is just one component of the overall neighbourhood package.

The significance of neighbourhood
What is clear for the women across the neighbourhoods in London is that becoming a mother means becoming more locally oriented. For many, especially in the inner London neighbourhoods, it means actually seeing and experiencing the neighbourhood for the first time.
because I was doing a demanding job, frankly Monday to Friday I could have lived absolutely anywhere and it would have been irrelevant you know to my day-to-day life ... having a baby is the first time really, since I was a child, that I've spent much time actually where I live um, and it's been terrific actually. (Fiona, gentrified neighbourhood, Balham, London) I fit into the community as a mum really. (Ella, gentrifying neighbourhood, Peckham, London) The Amsterdam parents stress the major change in their lives in terms of where they go to socialise and with whom. For most parents the neighbourhood becomes more important as a consumption space and a place where spare time is spent. While in Amsterdam this use and sociability of the neighbourhood is relatively even between male and female parents, in London it tends to be expressed through female respondents' experiences of motherhood. It comprises a growing familiarity with the neighbourhood and social networks (such as pre and post natal meetings). There is an increased recognition and appreciation of public services such as municipal swimming pools, parks and pre-school provision. It also provides the demand for retail and leisure infrastructure to cater to the needs of the middle-class mums with small children. These include baby clothes shops, boutique toy shops and, crucially, cafes in which middle-class mothers can meet to socialise. In London all these elements give middle-class mothers and their consumption practices a very strong presence in the public space of these neighbourhoods. It is a presence that is highly problematised in the accounts of many of the respondents in the two inner urban neighbourhoods studied. The problematisation of middle-class mothers in urban space is expressed through the discourse on the 'yummy mummy'. They don't stay at home, because either the builder or the cleaner is in there, they colonise the caf es. I went into Starbucks yesterday and counted the number of seats that were taken by three women with their children, nine seats altogether and that left about five seats for the rest of the Starbucks customers. (Catherine, gentrified neighbourhood, Balham, London) Well I suppose that, yeah. I suppose it is that, I am not meaning it really to be disparaging, I only, it's only disparaging when I want to be in there quietly and it'sf u l l of yummy mummies! [laughs]. So it is predominantly white middlethe people with money, will go therenot just white, not totally. (Antonia, gentrifying neighbourhood, Peckham, London) It is interesting to note that most of the discussions of yummy mummies came from other female respondents (some of them also parents). It connotes professional middle-class mothers who are affluent and mostly fulltime mums, but also well groomed and expensively dressed. This is a complex attribution because all the female parent interviewees also talked about the importance of social networks and sociability in the neighbourhood when dealing with the demands of bringing up small children. The idea of the yummy mummy is further evidence of the problematisation and emotionally charged nature of middle-class motherhood, a certain version of which becomes particularly visible in the context of socially mixed inner urban neighbourhood. It also represents the embodied co-constitution of gender and class in urban space. It is evident from the analysis of the London neighbourhoods that certain neighbourhood spaces are strongly feminised and associated with young motherhood in highly visible ways (the yummy mummies in Balham) or normalised as naturalised in the case of the commuter villages, and the suburb to some extent. The significance of female parents to school and village activity is obvious, it is a place where 'everything is done by mums'.
In contrast with the London context, the inner Amsterdam neighbourhoods were not important just for women. In the Amsterdam sample 'daddy days' are almost equally common as 'mummy days' and so the dominant presence of middle-class mums is not as striking as in many of the London neighbourhoods. There is an analogous symbol of middle-class parenting (in comparison with the 'yummy mummy'), that of 'cargo bike parents', who transport their children on bicycles with large boxes in the front in which the children sit. Interestingly, the cargo bike symbolises not just middle-class identity via material consumption, but also via the gender meaning it carries. The cargo bike has also become a symbol of the urban egalitarian parent, that is, the mother with a career and the father with serious caring tasks. The presence of this bike has become a marker of family gentrification and area status. Of course other -less symmetrically organisedfamilies also live in these areas, which still give these neighbourhoods a gender-bias. Yet, in many middleclass family areas the consumption infrastructure is geared to these middle-class dual parenting families rather than to mums per se.
A further contrast between female parents in Amsterdam and London is the degree to which the stresses of parenthood and its psychological impacts weigh quite heavily in the women's accounts from the London neighbourhoods. This is a strongly gendered experience (there was no such discussion from the male parents or in reference to them). The neighbourhood is a crucial site, as either a source of psychological support, or in compounding the experience of disconnection in the experience of 'being a mum'. Some of this is experienced as the initial contrast between a fulltime career and the transition to parenthood with parttime work or no formal employment: for a lot of women who've had a career, for them to say, I've no career, and that's it [laughs] what the hell am I gonna do, it would be horrible, becausebecause you know, you do lose a lot of your own identity, inin rearing children andand not doing anything else in yourself, and you know it can be very isolating as well. (Rebecca, gentrified neighbourhood, Balham, London) Clara: Lots of women are on anti-depressants around here. Another dimension of this is adjustment to the neighbourhood, which is experienced ambiguouslyranging from positive ascriptions of sociability and feelings of belonging, through to sociability being understood as necessary to counter feelings of dislocation and disconnection. I suppose this place really changed for us when we did have children; brilliant for bringing up children, so then I started meeting lots of people in this area ... Um, I guess, you know, having a child and being at home full-time I was determined that I needed to meet people otherwise I'dg o crazy. (Clare, suburb, Berrylands, London)

Discussion and conclusion
In this comparative study of London and Amsterdam we have demonstrated that the impact of parenthood is strongly gendered across professional middle-class groupings. By conceptualising parenthood as a field, with new rules and stakes, the changing practices of the middle classes can be seen as a reponse to the changing relationships between their social worlds (fields) and their class and gender dispositions (habitus). Parenthood is a widely experienced pinch point that raises the stakes in social reproduction (especially in the educational field). Parenthood as a field opens the habitus to a range of new policy influences, not least in educational policy, but also in parent/carer taxation regimes; workplace legislation, child care provision and cost. Parenthood thus has a profound effect on reconfiguring a number of related fields as well as being a site of social struggle in itself. How people respond to this transition, and how their practices change depends on the way in which the habitus becomes articulated in the new field of parenthood. It has been convincingly argued that class dispositions, rooted in the habitus, affect the way in which parenthood impacts on work, schooling and residential practices. In this paper we have argued that gender may play an even more important role and that the interaction of class and gender is crucial for understanding how the middle classes reproduce.
In Bourdieu's original formulation (Bourdieu 2001) gender was secondary to class and a hidden form of social stratification. In response, feminist theorists have variously argued that gender is a specific form of capital, is a form of symbolic violence against women and an experience of discontinuity or dissonance across fields for women (with its opportunities for rearticulation) (Adkins and Skeggs 2004;Huppatz 2009;McLeod 2005;McNay 1999;Reay 2000;Skeggs 1997). In the field of parenthood, we argue, the disruptive forces of a wider range of fields are felt over a short period of time. The relationship between habitus and field is disrupted by the emergence of the new field and its impacts on other fields (occupational careers; changing housing needs for example) and by opening up new areas (such as child care and education), which are subject to national institutional and public policy arrangements. It has been argued that the cumulative impacts of a socially differentiated range of fields (each with their own rules of the game) is strongly gendered and a source of women's uneven experience of degrees of autonomy and inequality (McLeod 2005;McNay 1999). Parenthood is a phase of instability, producing new field relations across a range of fields and institutional influences in which gender relations become critical. The field of parenthood thus rearticulates gender as the critical social process, of primary strategic importance (rather than secondary as Bourdieu assumed) in that it crosses social fields and draws in (often gendered) institutional arrangements. Furthermore the differentiated re-articulation (or 'refraction'; McNay 1999) of the habitus in response to changing field relations is strongly gendered. The Amsterdam and London comparison has shown how these fields are negotiated and how institutional influences and spatial (locational) adjustments are resolved differently in different contexts, resulting in varying time-space trajectories of the middle classes. The way that habitus is re-articulated to the changed/new fields is thus evident practically (the 4-4 egalitarian arrangement or the uneven gender split in Amsterdam and London respectively), but also symbolically (in the contrasting figures of the yummy mummy and cargo bike parents). These gender relations mean that being middle class is done differently in those gentrified neighbourhoods of Amsterdam and London and that the mechanism of class re-articulation (prompted by entering the field of parenthood) is through differing gender relations. The way that class practices are articulated in these neighbourhoods varies along gender lines resulting in distinctive neighbourhood milieuone that emphasises family gentrification (Karsten 2003), the other that singles out the figure of the middle-class professional woman as a conspicuous symbolic figure in urban spacethe 'yummy mummy'.
As our discussion of the gender divisions of labour and the relationship with formal employment indicates, employment practices and workplace cultures as well as social welfare settlements in different national states thus have a significant influence on how gender and class are reproduced. The rules of these fields, which are largely 'set' by the national legislation, strongly inform how becoming a parent impacts on the position in fields of employment. We have demonstrated how the transition to parenthood has a different impact on certain fractions of the middle classes in London and Amsterdam. Where there was evidence of more equitable childcare sharing in the Amsterdam case (with lower childcare costs and more family friendly employment practices) in London the childcare regime was consistently strongly divided along gender linesregardless of whether it was in urban, suburban or semi-rural neighbourhoods. In the London case inflexible work regimes and high childcare costs mean that a traditional gender norm around full-time male employment is reinforced.
In spite of the fact that the different types of neighbourhood encompass different middle-class fractions of the larger professional/managerial middle class grouping, the gender roles, especially in relation to childcare, are surprisingly uniform across them. This partly contradicts the idea that gentrifiers have more egalitarian gender practices, as described in literature on gender and gentrification (Bondi 1999;Karsten 2007;Warde 1991). It could be argued that some of the egalitarian gender norms among gentrifiers cannot be sustained in the institutional and spatial context of London. This is linked to workplace norms and legislation that are incompatible with child rearing and expensive child care but also to the increasing pressures of the London housing market, which involve trade-offs between fields of housing, employment, consumption and schooling. The Amsterdam case provides some contrast to this with inner urban gentrifiers negotiating a more egalitarian division of labour that stands in contrast to the rest of the population of parents more generally in the Netherlands and even within the middle classes themselves (see also Boterman and Karsten 2014). Here our findings resemble those of Bondi (1991Bondi ( 1999 and are also reminiscent of the discussion between Butler and Hamnett (1994) and Warde (1991) in which Butler and Hamnett argue that egalitarian gender practices are part of the work and life style of the urban new middle class. It seems that more equitable gender practices in the field of parenthood in Amsterdam are part of a distinctive cultural repertoire of a certain fraction of the middle classes, which can be sustained due to the relatively favourable child-care facilities and workplace legislation, but perhaps also by the 'human scale' of the city: short commuting distances and bicycle culture. The transition to parenthood does not involve the same radical trade-offs that we found in the London study.
The neighbourhood elements of time-space trajectories emphasise the significance and specificity of urban space in terms of the mix of neighbourhoods on offer to match the changing relations of habitus to field. In the past the suite of neighbourhood types that different cities offer and their effects on habitus have been discussed in terms of trade-offs between different forms of capital (economic, social, cultural) and the degree to which there can be considered to be a 'metropolitan habitus' (Bridge 2006;Butler and Robson 2003a). In the current study it is the conjunction of the 'neighbourhood offer' in the two cities with the constraints that come with the field of parenting that is significant. In the Amsterdam case this gives a cultural distinctiveness to certain family gentrification neighbourhoods, while in London seemingly different neighbourhood types mask similar gendered practices in terms of class reproduction.
The findings from Amsterdam could also be linked to other studies that have proposed to integrate cultural dimensions of the habitus with the symbolic meaning of how femininity and masculinity, in this case motherhood and fatherhood, are practised (Huppatz 2009). The egalitarian parent, also typified by specific consumption practices such as driving a cargo bike, is a strong symbolic marker of class via gender practices. The uniform gendered division of labour in London means that these aspects of gender practices are little used for distinction. Nonetheless, different ways to be a mum (i.e. yummy mummies) are also ways in which femininity is used a symbolic marker of class.
This comparative analysis further problematises the relationship between habitus and field. A strong stream of urban research has seen the relationship between habitus and field largely in culturalist terms in which adjustment of habitus to field is, in part, spatial, through a form of neighbourhood matching of habitus and field in forms of 'elective belonging' (Savage et al. 2005). The emphasis lies on cultural dimensions of class while space is quite passive in this conception. Although cultural dimensions do clearly emerge from our analysis, we nevertheless point to the significance both of institutions in the wider political economy, but also to a more activist view of the way that space is worked through in gender and class terms in the neighbourhood contexts. Both these institutional and spatial realms emphasise the significance of gender relations in class reproduction. The central institutional mechanism for class reproduction for Bourdieu is the education system (Bourdieu and Passeron 1977). In the present study we have seen that institutional mechanism through the lens of parenting and this has pointed out the gendered aspects of the effort in reproduction via education (playground grapevines, liaison with teachers, school choice, bearing the 'costs' of education-driven residential moves, for example). We have also suggested how these elements are also affected by the relationships between institutions in different state welfare and social policy settlements (especially in workplace flexibility and childcare regimes and costs). This suggests that we need to acknowledge the institutional mix and the part played by different political economies in class and gender reproduction.
On the other hand we suggest a greater significance for space and the gendered way that space is 'worked' in parenting, from simple physical presence in the neighbourhood to the importance of child-based social networks, use of and involvement in local services (schools, parks, social clubs), and the management of social mix and use of public space. Again these workings of neighbourhood space are highly gendered for the most part (except where the institutional and urban spatial arrangements coincide with particular cultural fractions of the urban middle class). Thus parenting as a field disrupts established habitus-field relations but also points to intra-household gendered differences (and cross-neighbourhood gendered continuities) in the process of attempting to reconcile habitus and field and the spatial effects of these efforts in terms of differentiated forms of 'belonging'.
The original literature on gender and gentrification pointed out the significance of the increase in women professionals in the labour market (Warde 1991), of urban space for managing different tasks for women marginal gentrifiers (Rose 1984), of urban context in understanding gender and class constitution in gentrification (Bondi 1991) and of seeing more emancipatory gender practices within the confines of a relatively privileged (new urban) middle-class fraction (Butler and Hamnett 1994). This study endorses these earlier debates on the significance of female participation and of the urban context in the changing middle classes, but suggests that the more egalitarian gender practices associated with the new urban middle classes are themselves conditioned by the larger political economy (of welfare regimes in this case), by the ways that competing fields can be reconciled at the city scale and by the role of neighbourhood as a site of field management and class practices at the local scale. Rather than a simple ascription of class identity to neighbourhood type (as in ideas of elective belonging), we need to acknowledge both the particular mix of neighbourhoods at the city scale as well as everyday practices across neighbourhoods and within them in understanding social reproduction and identity. Urban space acts as both a constraint (the metropolitan mix of neighbourhoods) and as an active site in the working through of classed and gendered practices. The adjustment of class habitus to field puts gender as the critical mechanism in different patterns of class re-constitution. A comparative analysis of the field of parenthood in relation to gender and class habitus points to the critical intersections of political economy, locational amenity and everyday practices as well providing an integrative theoretical approach to the changing relations between gender, class and space.
1 The Equal Treatment (Working Hours) Act (Wet verbod op onderscheid naar arbeidsduur, WOA) became law in the Netherlands in 1996. This Act prohibits an employer from discriminating between full-time and part-time employees, unless there is an objective justification for doing so. The underlying principle of the Act is that parttime work is equivalent to full-time work. 2 Public policy in the UK was for a long time distinctly gendered, first based on a male breadwinner/female homecarer model, which then shifted to a male breadwinner/ female part-time carer model (Sciarra et al. 2004).
Between 1975 and 1995 those working less than eight hours per week were disqualified from many statutory rights, while those working between eight and 16 hours had to demonstrate five years of continuous employment to qualify for these rights. During these years, in line with the government policy, part-timers were usually given worse terms and conditions of employment than full-time workers, generally being the first ones to be made redundant. Part-time workers were therefore crowded into the lowwage sectors of the economy.