‘GOOD’ MOTHERHOOD:
COMPATIBLE OR
CONFLICTING
WITH PAID LABOUR?
The Gendered Moral Rationalities of South African Mothers
concerning the Relation between ‘Good’ Motherhood and Paid Work
Master Thesis - Research Master Social Sciences
Marthe Scheepens (10556184)
August 5
th, 2019
Abstract
In this research, the relationship between ‘good’ South African motherhood and paid work is examined. To find out whether South African mothers understand ‘good’ motherhood as compatible or conflicting with paid labour, this research uses Duncan and Edwards’ (1999) concept of gendered moral rationalities to gain insight into South African mothers’ conceptualisation of ‘good’ motherhood. Three ‘good’ motherhood rationalities are distinguished: (1) attached motherhood, (2) shared motherhood, and (3) strong motherhood. In addition, Duncan and Edwards’ (1999) rationality framework is used as a starting point to grasp mothers’ moral understanding concerning the division of paid and unpaid work. The three ideal types as determined by the authors are also detected in this research (although with a small shift): (1) flexible primarily mother, (2) primarily worker, and (3) mother/worker integral. By looking at the reflected combinations of ‘good’ motherhood rationalities and rationalities concerning right’ division of labour, this research distinguishes four different rationality combinations that seem mainly socially patterned on the basis of economic class. The understanding of ‘good’ motherhood and paid work as conflicting is linked to middle-class mothers, while the understanding of ‘good’ motherhood and paid work as compatible is related to economic disadvantaged (black or coloured) mothers.
Contents
Abstract
Chapter 1: Introduction 4
Chapter 2: Theoretical framework 8
2.1. Shortcomings of neoclassical theory and structuralism 8
2.2. Gendered moral rationalities 10
2.3. Gendered moral rationality framework 11
2.4. Shortcomings gendered moral rationality framework 13
Chapter 3: South African Context 14
3.1. South Africa’s diverse society 14
3.2. South African social groups 15
3.2.1. Black South African mothers 15
3.2.2. White South African mothers 16
3.2.3. Asian South African mothers 17
3.2.4 Coloured South African mothers 17
3.3. Research side 18
3.4. Contemporary South African motherhood 18
Chapter 4: Methodology 20
4.1. Participant sampling & Recruitment 20
4.2. Participant group 21
4.3. Data collection & Operationalisation 23
4.4. Analysis of the data 24
4.5. Ethical considerations 25
4.6. Methodological limitations 25
Chapter 5: Results 28
5.1. Conceptualisation of ‘good’ motherhood 28
5.1.1. Attached motherhood rationality 28
5.1.2. Shared motherhood rationality 32
5.1.3. Strong motherhood rationality 35
5.2. Rationalities concerning the ‘right’ division of labour 38
5.2.1. Primarily mother rationality 38
5.2.1.1. Flexible mother rationality 39
5.2.2. Primarily worker rationality 41
5.2.3. Mother/worker integral rationality 42 5.3. Relation ‘good’ motherhood and paid work: compatible or conflicting? 44 5.3.1. Attached motherhood & flexible primarily mother 45 5.3.2. Attached motherhood & primarily worker 46 5.3.3. Shared motherhood & primarily worker 47 5.3.4. Strong motherhood & mother/worker integral 47
Chapter 6: Conclusion 49
6.1. Research implications 50
1. Introduction
From the moment South Africa became a democracy in 1994, its government has striven to overcome the legacy of the apartheid system by eliminating poverty and reducing inequality. The economic inclusion of women is perceived as crucial for achieving these goals because it enables the country to become a stronger player in the global market place. In addition, investing in women’s economic participation is a matter of gender equality and support women’s human rights (DPRU, 2015, p. 7). Although great progress has been made in improving female labour participation since democracy was established, the South African government considers it necessary that more women enter the labour market. According to The Report on the Status of Women in South Africa Economy (2015), women’s economic participation can be stimulated by the removal of structural barriers. However, the relatively low participation of women in the labour market and the variation of this participation among different South African socioeconomic groups is only partly related to economic and structural reasons. Women do not simply react to policies or economic constraints (Pfau-Effinger, 1998, p. 147-148). The South African government stresses that gendered attitudes, assumptions, and traditions play an important role in relation to labour participation (DPRU, 2015, p. 9). The Report on the Status of Women in South Africa Economy (2015) concluded that gendered ‘norms are powerful in shaping behaviour’ and are crucial to the opportunities and constraints that men and women experience in a labour market (p. 145).
Normative ideas about womanhood are strongly related to moral notions of motherhood. A significant proportion of South African women who do not participate in the labour market are mothers whose responsibilities include unpaid household work and childcare activities (DPRU, 2015, p. 110). To develop policies and (attitudinal) interventions that encourage women’s paid labour participation, it is important to understand mother’s normative understandings with regard to motherhood. A motherhood ideology describes the social norms, beliefs, and values of the societal understanding of ‘good’ mothering (Hattery, 2001; Kruger, 2006; Pheonix & Woollett, 1991). Ideologies shape and are shaped by historical and social conditions and can have a powerful effect on individuals, societies, and sub-societal groups (Johnston & Swanson, 2006, p. 510). They inform and (de)legitimise practices but are not the same as a practice because individuals have agency to accept or reject the dominant ideals (Walker, 1995). Insights into mothers’ understandings regarding ‘good’ motherhood and the relation to paid work are important for policymakers because “although structural resources matter, inherent cultural beliefs and practices exist [that] encourage some mothers, but discourages others, to work outside the home” (Dow, 2016, p. 1940). The current literature does not provide adequate insights into the conceptualisation of ‘good’ South African motherhood in relation to labour because it focuses primarily on Western (i.e., white middle-class) motherhood.
In the existing literature, the Western-based intensive motherhood ideology is often taken as the norm or contested standard (Hallstein, 2006, 106; Johnston & Swanson,
2006, p. 510; Murray & Finn, 2012 p. 44). This motherhood ideology refers to the stay-at-home mother who is primarily responsible for the household and the care of children. It is strongly linked to the capitalist division of the private and the public sectors: while the woman is at home to take care of the children, the man is engaged in paid work outside the home to provide income for the family (Katz-Wise, Priess & Hyde, 2010, p. 18-19). Current studies on motherhood repeatedly show the ambivalent experiences of working mothers that arise from the conflicting expectations of the logic of the intensive motherhood ideology and the market logic that focuses on profit, success, and personal achievement (Blair-Loy, 2003; Damaske, 2011; Gerson, 1985; Hays, 1996; Stone, 2007, Walzer, 1998; Wearing, 1984).
This conflict between the separate spheres, therefore, seems related to a Western conceptualisation of ‘good’ motherhood. Many scholars argue that the ideal of intensive mothering is influenced by characteristics of social status including race, class, sexual orientation, national origin, and immigrant experience. According to them, the ideology is related to Eurocentric and privileged white, upper middle-class women and that this dominant ideology is not applicable to all (sub)societies (Collins, 1991; Duncan & Edwards, 1999). As McMahon (1995) concisely explains, “motherhood is constructed as the expression of women’s natural, social, and moral identity—or, rather, the identity attributable to moral women, that is married white women” (p. 24). Black feminist scholars, like Collins (1994), are critical about the ethnocentric assumptions of motherhood as “natural, universal, and unchanging” (Jeannes & Shefer, 2004, p. 3) and argue that “mothering takes place within specific historical contexts framed by interlocking structures of race, class and gender” (p. 45).
The existing literature mainly reflects the image of culturally privileged white Western mothers within a nuclear family setting (O’barr, Pope & Wher, 1990, p. 2; Bhopal 1998, p. 485) and largely excludes motherhood constructions of those who lie outside the Western motherhood perspective such as non-white mothers or those who are poor (Liamputtiong, 2006, p. 27). Limited research on non-white mothers shows that motherhood and work is often more integrated and that the motherhood-worker dichotomy that is represented in the dominant literature needs to be questioned.
Mothers with different cultural and economic backgrounds seem to differ in their conceptualisations of ‘good’ motherhood and, thus, understand decisions concerning employment in a different way. While some mothers consider motherhood and paid work as easily compatible, others believe that ‘good’ mothering and paid labour are conflicting. To obtain more insights into labour considerations of mothers from different socioeconomic groups in a multiracial society, this research examined the relationship between ‘good’ South African motherhood and the ‘right’ division of labour. In particular, the following research question was asked: ‘How do South African mothers understand the relationship between ‘good’ motherhood and paid work?’
To answer this question, Duncan and Edwards’ (1999) concept of gendered moral rationalities was used to gain insights into the conceptualisation of ‘good’ motherhood
and the understandings concerning the ‘right’ division of labour. According to the authors, mothers’ decisions concerning paid work are based on more than a rational cost-benefit analysis. Instead, as also frequently mentioned in ‘The Report on the Status of Women in South Africa Economy’ (2015), gendered norms, attitudes, and assumptions also play an important role in explaining women’s labour force participation. Gendered moral rationalities focus on the normative aspects of motherhood and help understand how the conceptualisation of ‘good’ motherhood is related to decisions concerning employment. Duncan and Edwards (1999) stated that structural theories are insufficient in understanding individual work-family considerations because they cannot declare why mothers make different decisions when they live in the same period under the same rules. The authors argued in contrast that individuals are rational beings but question the neoclassical assumption of the “separate self”. The concept of gendered moral rationalities links agency with structure by showing that mothers’ individual moral rationalities are socially negotiated and culturally patterned. Hence, the concept is particularly helpful in the context of South Africa, in which mothers of different social groups differ in their employment decisions. Furthermore, Duncan and Edwards (1999) proposed three ideal types of moral gendered rationalities that explain variations of mothers’ employment decisions. This research used Duncan and Edwards’ (1999) framework as a starting point to gather insights into the relationship between ‘good’ motherhood and employment. However, it also demonstrated that their framework is not extensive enough to understand whether mothers consider ‘good’ motherhood and paid work conflicting. Mothers who consider the uptake of paid work as morally ‘right’ can differ in their understandings concerning ‘good’ motherhood. This research showed that insights into the conceptualisation of ‘good’ motherhood as separate from the relationship to work is important to better understand whether the relationship between motherhood and paid work is understand as compatible or conflicting by South African mothers. Paid work in this research refers to a range of substantial part-time work to full-time employment.
By examining the relationships between ‘good’ motherhood and paid work, this research aimed to enrich the existing literature that assumes Western middle-class motherhood as the standard (Liamputtong, 2006, p. 27). Evidence of variations in mothers’ conceptualisations of ‘good’ motherhood by social group would provide valuable insights about the normative structures that influence women’s labour participation. Therefore, by better understanding these influencing factors, the results of this research can contribute to the goal of encouraging economic labour participation in South African mothers.
In Chapter 2, Duncan and Edwards’ (1999) concept of gendered moral rationality and the authors’ rationality framework of three ideal types is explained in further detail. Chapter 3 focuses on mothers’ labour participation in the South African context. Chapter 4 reports the methodology of this research. Chapter 5 presents the results organized
into three sections: (1) mothers’ conceptualisation of ‘good’ motherhood, (2) mothers’ understanding of the ‘right’ division of labour, and (3) the relationship between ‘good’ motherhood and the ‘right’ division of labour. In all sections, the socioeconomic context in which rationalities are negotiated are stressed. In the concluding chapter, it is demonstrated how different social-economic groups of South African mothers understand the relation between ‘good motherhood and paid work. In addition, the theoretical and social implications of the research are explained, and suggestions for further research described.
2. Theoretical Framework
To stimulate South African mothers’ labour participation, it is important to gain an insight into what South African mothers consider ‘good’ motherhood and see as ‘right’ choices regarding paid and unpaid work. This research makes use of Duncan and Edwards’ (1997) concept of gendered moral rationalities. This chapter explains why this concept is particularly useful for examining the normative work and family considerations of South African mothers. First, the inaccuracy of neoclassical theories is explained using two major theoretical approaches on work and motherhood decisions: (1) the preference theory and (2) new household economics. Next, the shortcomings of structuralism are presented for explaining mothers’ work and family considerations (section 2.1.). Then, the gendered moral rationalities are positioned and explained (section 2.2.). Furthermore, Duncan and Edwards (1999) rationality framework is examined in further detail (section 2.3.), and its limitations are described (section 2.4.).
2.1. Shortcomings of neoclassical theory and structuralism
To understand mother’s moral considerations regarding unpaid and paid work, Duncan and Edwards (1997) developed the concept of gendered moral rationalities in reaction to two neo-economic theories: (1) the preference theory and (2) new household economics. These theories explain work and family considerations on the basis of rational economic choice and individualisation. Economic choice theories reflect the idea of individuals as economic agents (homo economicus) with fixed preferences that are independent from others and past behaviour and who make cost-benefit analyses to maximise their utility. Utility refers to financial advantages but also includes ‘taste for leisure’ (Duncan & Edwards, 1997, p. 33). The individualisation approach is related to Gidden’s (1992) idea that ‘the self’ has become a reflexive project that is no longer determined by social positions (May, 2011, p. 364). As a result, life trajectories are no longer bound by traditional structures and individuals are able to individually negotiate unconstrained and unforced choices (Duncan & Irwin, 2004, p. 391).
The first neo-economic theory that Duncan and Edwards drew upon was the preference theory of Hakim (1996, 1988, 2000, 2002), which explains women’s choices related to market work and family work on the basis of ‘individual preferences’. According to Hakim, individuals act rationally to maximise their desires and have stable preferences (1996, p. 13). Structural processes and social characteristics, such as class and gender, impose only minor constraints on women and men, or in some cases none at all. “There are no major constraints limiting choice or forcing choice in particular directions” (Hakim, 2000, p. 18; McRae, 2003, p. 328). According to the theory, the heterogeneity in preferences explains differences in work and family positions between women. Hakim identified three patterns of women’s preference concerning work and family: work-centred women who have serious careers; women who stay at home and consider their family the main priority in life; and adaptive women who are not career focused and (wish to) combine work and family (McRae, 2003, p. 218).
The second theory to which Duncan and Edwards reacted is the new household economics of Becker (1981, 1996). This theory is also based on neo-classic economic
assumptions but extends them to gendered division of labour. In contrast to Hakim’s preference theory, the household—not the individual—is considered as the most important unit for utility maximisation. According to Becker, spouses are mutual dependent and optimise the division of labour by specialising into functions that capitalise their human capital best; women specialise in household and care activities and men in market work (Duncan & Edwards, 1997, p. 34-35). Households involve the rational economic trading of different and complementary specialisms and is thus both a unit of production and a unit of consumption. Becker argues that the gendered division of labour is a result of both biological differences and socialisation and assumes that preferences are fixed (Katz, 1997, p. 27).
Both the preference theory (Hakim, 1996, 1988, 2000, 2002) and new household economic model (Becker, 1981, 1996) are controversial because they mainly neglect the influences of structural processes on decision-making and social action. Many researchers have argued against the neoliberal idea of individualisation, believing that considerations about paid and unpaid labour are always shaped by wider economic opportunities and cultural values. ‘The self’ and the society cannot be seen as two separate spheres (May, 2011, p. 365-366; McRae, 2003, p. 328-329; Thwaites, 2017, p. 56). McRae (2003) argued that structural and normative constraints can affect women’s considerations concerning work and family. Structural constraints include not only job availability and the cost and availability of childcare but also differences in social background that are related to educational opportunities, teenage pregnancies, health problems, and culture (McRae, 2003, p. 329). Normative constraints include “women’s own identities—their ‘inner voices’—gender relations in the family, and husband/partner’s attitudes” (p. 329). According to McRae (2003), structural and normative constraints can intertwine with each other and affect decisions concerning work and family (p. 329).
Duncan, Edwards, & colleagues (2003) argued that only structural constraints and opportunities, including state policies, do not alone explain mothers’ decisions about paid work and family responsibilities (Duncan, Edwards, Reynolds & Alldred, 2003, p. 310). According to them, structural constraints do not explain why mothers make different employment decisions even though they live in the same period under the same rules. Duncan and Edwards argued that social research often includes social context into debates but frequently represents individuals as agents who simply respond to their context; by contrast, the authors assert that women are not “victims of external material constraints” (Lister, 1999, p. 237).
As stressed by the South African government, gendered ‘attitudes, assumptions, and traditions’ play an important role in relation to labour force participation (DPRU, 2015, p. 9). Normative constraints are essential in understanding women’s considerations concerning paid and unpaid work. Duncan and Edwards (1999) argued that the power relation of gender is an important normative constraint, even though it is
often unseen because women internalise less powerful positions within their ‘preferences’ (p. 34).
In addition, the authors (1999) stressed that economic theories that focus on human capital (e.g., education), individual resources (e.g., income), and constraints (e.g., number of children) lack insight into the agents’ collectively negotiated understanding of the situations. Duncan and Edwards (1997) maintained that it is important to focus on social relations and moral understandings in explaining why women pursue paid work (p. 40). Thus, in reaction to the individualisation approach and structural theories, Duncan and Edwards (1997) developed the concept of gendered moral rationalities as a ‘middle element’ to provide insights into mothers’ socially negotiated normative and moral understandings about the division of paid and unpaid labour.
2.2. Gendered moral rationalities
To understand mother’s moral work and family considerations, Duncan and Edwards (1997) developed the concept of gendered moral rationalities. Gendered moral rationalities are cognitive schemas that provide a guiding framework about what a ‘good’ mother should do (Duncan, Edwards, Reynolds & Alldred, 2003, p. 313). According to Duncan and Edwards (1997), “it is gendered moral rationalities—social and cultural collective understandings about what is best, and morally right, for men as well for women—that provide a basis for individuals’ economic decision-making within families” (p. 35). In addition, these moralities are highly gendered. Duncan and Edwards’ concept of gendered moral rationalities draws upon contemporary feminist theories that argue that women are not a homogeneous group. “Gender is a social division which intersects with other sources of social division” (Lister, 1999, p. 236).
For the concept of gendered moral rationalities, Duncan and Edwards (1997) adopted the common belief of economic theories that individuals are rational beings but questioned the neoclassical assumption of the “separate self”. Duncan and Edwards emphasised that “people are indeed highly rational (given the imperfect information they hold) and they do weigh up the costs and benefits of alternative courses of action— but both what is rational, and what constitutes a cost or a benefit, are defined in collective moral and social terms, not simply in individual utility terms” (Duncan & Edwards, 1997, p. 56). Duncan and Edwards argued that economic rationalities like human capital, individual resources, and constraints are subordinated causes to moral understandings and only become important when social rationalities are established (Duncan & Edwards, 1997, p. 40).
In addition, the authors stressed that preferences (to use the economic term) concerning family and employment are socially negotiated, sustained, modified, and changed within particular contexts. Preferences are not fixed and cannot simply be found in individuals but arise out of social negotiation (Duncan & Edwards, 1997, p. 56). Moralities operate “at the interface of a national gender order—that is, the set of knowledges, values and ideals institutionalised in national family social polices—and local gender cultures—that
is, the socially sanctioned practices for men and women that exist in different neighbourhoods, among different social groups” (Ralph, 2016, p. 67). As a result, the gendered moral rationalities concerning labour market positions vary substantially between societies and, in particular, social groups, such as different race, religious, and class groups. Because gendered moralities are socially negotiated and culturally patterned, they can in turn provide information about gender structures within a particular context (Duncan & Edwards, 1997, p. 56). “Thus, while gendered moral rationalities are a product of human-collective-agency, rather than being a property of social structure over which individuals have no control, there is also a sense in which they become structural features. As socially patterned moral guidelines, they both constrain and facilitate certain courses of action for particular social groups” (Duncan & Edwards, 1999, p. 129).
2.3. Gendered moral rationality framework
Duncan and Edwards (1999) identified three main types of moral rationalities about mothering in relation to employment in their research about lone mothers with different racialised and economic backgrounds in the UK. These rationalities concerning the ‘right’ division of labour are linked to mothers’ rationalities about childcare that are based on (1) their understanding of their children’s needs, (2) their own needs, and (3) the balance of these needs (Duncan, 2005, p. 57). According to Duncan and Edwards (1999), motherhood is a social relationship: for there to be mothers, there must be children. Motherhood as a relationship is constituted in terms of children’s needs, which invokes a moral urge that these needs must be met. The moral compulsion to be responsible for (and meet) children’s needs is vested primarily in mothers: “’good’ mothering is meeting children’s needs” (p. 119). Duncan and Edwards (1999) argued that ‘the overwhelming majority’ of participating mothers feel a general responsibility towards their children; however, the way that mothers understand their children’s needs is not unidimensional. Mothers differ in how they think their responsibility towards their children should be discharged. These differences in the understanding of children’s needs yields varied ideas about the compatibility of motherhood and paid work (Duncan & Edwards, 1999, p. 117). Hence, variations in the understanding of ‘children’s needs’ result in different courses of action in relation to the uptake of paid work. In addition, different aspects of children’s needs can result in different, and sometimes conflicting, courses of action (Duncan & Edwards, 1999, p. 131). Furthermore, Duncan and Edwards (1999) argued that the understanding of ‘children needs’ is formulated within particular socioeconomic and cultural contexts, making the understanding of ‘good’ motherhood collectively negotiated and socially patterned. Duncan and Edwards (1999) focused their framework of ideal rationalities on the conceptualisation of ‘good’ motherhood in relation to paid work. The first rationality is the ‘primarily mother’. The ‘primarily mother’ gives moral primacy to caring for their own children over any form of employment. Mothers with this rationality prefer the role of traditional homemaker and do not want a long part-time or full-time paid job.
The second rationality, the ‘primarily worker’, embraces a mother identity and childcare responsibilities but gives moral primacy to paid employment (Hagelskamp, Hughes, Yoshikawa & Chaudry, 2011, p. 339). The mothers in this category stress that employment positively affects “the self” and attribute feelings of pride and commitment to employment. The identity as employee remains separate from the identity as a mother. According to Duncan and Edwards, mothers with ‘primarily worker’ rationalities experience work-family conflicts and experience guilt when employment results in the abdication of motherhood responsibilities (Hagelskamp et al., 2011, p. 339). ‘Women with higher levels of education and those who were married tended to justify their involvement in paid work by avowing their passion for their work whereas less educated women and single mothers stressed the importance of women’s economic independence’ (Hagelskamp et al., 2011, p. 339).
The third rationality Duncan and Edwards (1999) defined is the ‘mother/worker integral’ rationality. Mothers who reflect this rationality try to combine paid employment with the care for children and see employment as part of their responsibility towards their children. The role of breadwinner is incorporated in the identity as a mother. These mothers see it as their moral responsibility to be financially self-reliant and prefer to be a role model that advocates the importance of ‘hard work’ for money.
Although not all mothers fit perfectly within this framework, Duncan and Edwards (1999) argued that it challenges the dichotomy of mother versus worker that is dominant within the existing literature (Lister, 1999, p. 236). Previous research has often presented mothering and employment as two contrasting identities. Researchers asserted that women can hold the two identities simultaneously, but it results in feelings of conflict and guilt (Blair-Loy, 2003; Damaske, 2011; Gerson, 1985; Hays, 1996; Stone, 2007, Walzer, 1998; Wearing, 1984). Duncan and Edwards stated that the mother versus worker dichotomy reflects an ethnocentric perspective and represents only the experiences of white Western women. “Mother and worker constitute two prominent social role identity options for most women in many parts of the world. The meaning of these roles, and their associations with each other vary across social and cultural contexts, and within individuals” (Hagelskamp et al., 2011, p. 337-338). In their framework, Duncan and Edwards showed that black mothers primarily refer to a mother/worker integral rationality.
According to Duncan and Edwards (1997), ethnic differences are crucial in declaring employment patterns (p. 36). Their data shows that African-Caribbean mothers and black and African mothers show higher employment rates and are significantly more focused on full-time work than white single mothers. Indian and Pakistani mothers have the lowest employment rates, but those who do work focus on full-time employment (Duncan & Edwards, 1997, p. 47). This pattern also remains after correction for ‘human capital’ and ‘individual constraints’ (Lister, 1999, p. 236).
2.4. Shortcomings gendered moral rationality framework
Although Duncan and Edwards introduced a useful concept and framework of gendered moral rationalities, their theory also has some limitations. One of the major problems is that the authors try to unravel economic and moral rationalities. Although Duncan, Edwards, and colleagues (2003) did not deny that cost-benefit calculations, personal income, and promotions can have an impact on the decision-making process concerning work and family, they argued that these economic rationalities are secondary to moral rationalities (Duncan, Edwards, Reynolds & Alldred, 2003, p. 311). “There is a danger in Duncan and Edwards' exposition of overstatement of their case. It is one thing to challenge the grip of 'economic rational man' and to assert a place for gendered moral rationalities in the decisions that lone mothers make about employment. It is another to assert that the latter are a priori primary determinants in that decision making” (Lister, 1999, p. 238).
The data of Duncan and Edwards’ research (1999) indicate that the correlations between mothers’ gendered moral rationalities and economic rationalities and material context are complex. In addition, Hagelskamp, Hughes, Yoshikawa and Chaudry (2011), who studied the work and family considerations of poor migrant mothers, showed the interplay between economic and moral rationalities. The authors refer to mothers who lack social and economic resources and see employment as necessary but also have gender role expectations, cultural norms and traditional family relationships that conflict with their provision of income. Economic rationalities seem to intervene more with moral rationalities than Duncan and Edwards suggested, so it is important to show the interplay between them.
In addition, Duncan and Edwards focused their gendered moral rationality framework on mothers’ understanding of the ‘right’ division of work. Although these understandings depend on mothers’ different conceptualisations concerning ‘good’ motherhood, the authors’ framework does not focus directly on how these two different types of rationalities relate. Duncan and Edwards framework (1999) remains especially unclear about the moral understandings of ‘primarily workers’ concerning the relationship between ‘good’ motherhood and labour. Thus, for this research, ‘good’ motherhood rationalities are disentangled from rationalities concerning the ‘right’ division of labour (although these are motivated by ‘good’ motherhood rationalities). By considering these rationalities as two separate types of rationalities, it is possible to focus on the relationship between ‘good’ motherhood and labour and gain further insights into why certain mothers understand ‘good’ motherhood and paid work as compatible, while others believe they are in conflict.
3. South African Context
3.1. South Africa’s diverse society
This research focused on the experiences of South African mothers. South African motherhood is particularly interesting because of the country’s complex history. As a result of colonialisation and the apartheid era, modern South Africa is a multiracial society distinguished historically by four different racialised groups. In this study, ‘black’ refers to people who are descendants of the original inhabitants of South Africa. During apartheid, these people were also often classified as ‘native’, ‘Bantu’, ‘non-white’, and ‘African’. ‘White’ refers to ‘Afrikaner’ people who are descendants of Europeans. ‘Coloured’ refers to individuals who are born of, or descended from, racially mixed parents. The fourth group consists of descendants of immigrants from Indian, China, or Middle Eastern countries, referred to as ‘Asian’, ‘Indian’, or ‘Cape Malay’, respectively (Posel, 2001, p. 51). In modern South Africa, the legacy of colonialisation and the apartheid era is still visible. The country has a very diverse population, almost 80% of which is racialised as black. The minority groups in the country are racialised as white (9.2%), coloured (9.0%), and Indian (2.6%) (Coovadia, Jewkes, Barron, Sanders & McInyre, 2009, p. 817).
Both nuclear and extended families are prevalent in this multiracial society. Statistics show that 39% of households are nuclear families while slightly less (36%) are extended (STATS SA, 2018, p. 9). In contrast to nuclear families, extended families feature kinship relations that are not solely based on ‘blood lines’, and the childcare is shared among family and community members (Sudarkasa, 2004, p. 2). Thus, mothering in (African) extended families is a collective, communal responsibility. The way motherhood is defined can be “affected and impacted by the type of family structure or kinship grouping in which it is lodged” (Sudarkasa, 2004, p. 2). “Whereas traditionally, motherhood encouraged women to become more economically independent in African conjugal and extended families, motherhood may encourage some women to become more economically dependent in Western nuclear family settings” (Sudarkasa, 2004, p. 9-10). Economic provision is an integral component of motherhood, and the private-public dichotomy is less relevant for women within extended family structures (Segura, 1994). Race and class are structural features that underlie the diversity of family norms (Ferree, 1990, p. 879).
Besides this link between the mixed racialised population and family structures, there are also huge income differences in South Africa. The country is ranked as the society with the starkest income inequalities in the world, scoring the highest Gini index (63.4) in 2014 and a 7.1 Palma ratio in 2015 (UNDP, 2018). The economic differences in the South African society are huge and correlate with different racialised groups. Inequality between racialised groups is a result of the apartheid system in which non-white inhabitants did not have access to most forms of capital (Orthofer, 2016, p. 20). Although much has changed since 1994, non-white households are on average still much poorer than white households. However, the differences within the black population exceeds the level of inequality for the overall population and is much higher than within
any other racialised group (Orthofer, 2016, p. 4). Besides race, gender is related to economic position. Female-headed households in South Africa have a 48% chance of being poor, which is way higher than households headed by men (28%) (Woolard, 2002, p. 3).
A mother’s economic situation is strongly, but not exclusively, related to employment (Woolard, 2002, p. 5). Although the female labour participation rate in South Africa rose to 60% since the mid-1990s, currently one-third of the labour force population is not economically active. This group can be considered as voluntarily at home (STAT SA, 2017, p. 1). Almost 30% of the female labour participants are involuntarily at home because of the high female unemployment rates in South Africa (STAT SA, 2019, p. 19). Vast differences exist between the different South African racialised groups in the proportion of women who are participating in the labour force and of those who stay at home.
3.2. South African social groups
As a result of the unique history of South Africa, the construction and experience of motherhood differs along racial and socioeconomic lines. To understand the normative conceptualisation of contemporary South African motherhood, the (historical) material and ideological differences of social groups must be taken into account. In this section, the four racialised groups are used to describe social differences. This is not to say that racial backgrounds are the main features to explain socioeconomic and cultural/normative differences between mothers of various social groups, but South African statistics resources use these four racialised groups to provide information.
3.2.1. Black South African mothers
The early twentieth century in South Africa is characterised by overwhelming migration of black men to urban areas, while many married African women were restricted to the rural reserves. During the 1930s, the main identity of African mothers became that of carers, and they depended economically on their migrant husbands (Moore, 2013, p. 153). This identity was also strongly grounded in political organisations and campaigns (Walker, 1995, p. 420). Urban mothers depended on men too and were also accused of destabilising African family life (Moore, 2013, p. 153). During the 1950s, marriage became important to ensure housing and economic stability, but in the 1970s, social housing became more accessible, and women could also engage in more varied types of work. During this period, many black African women divorced in the interest of empowerment. The number of female-headed households increased and the stigma attached to single motherhood declined. To protect children from violence and disruption as result of political upheaval, many mothers sent them to rural areas. In many rural areas, grandmothers became the primary carers of their grandchildren. This extended family structure also enabled women’s involvement in employment (Posel, Fairburn & Lund, 2006). Later, the extended family structure, in combination with increasing male unemployment, resulted in significant changes between South African men and women. “Organised, resilient, women-centred networks became central to how
women managed mothering activities. Motherhood was altered less by the absence of husbands than by the growing importance of other women” (Moore, 2013, p. 154). Towards the end of the apartheid era, urban houses became scarce. As a result, adolescents could not leave the family home and matriarchal, multi-generational households emerge. During this period, African women still considered marriage as an ideal but were sceptical about the implications in their circumstances. In the 1980s and 1990s, educational attainment and employment arose among African women, and motherhood “took place alongside formal employment, home ownership, scholarship, and informal relationships” (Moore, 2013, p. 154).
Today, both African men and women participate in the labour force. In line with the integrated motherhood ideology, statistics of 2005 show that the labour participation gap within the African population is relatively low (11.6%) compared to the white and Asian population. Moreover, between 1995–2005, the female labour force of African women increased by 17% (Van der Westhuizen, Goga & Oosthuizen, 2007, p. 10). In addition, the extended family structure (43%) was more prevalent than the nuclear family structure (41%) within the African population in 2005. Notably, the number of African single-parent households were also relatively high (17%) (Amoateng, Heaton & Kalule-Sabiti, 2007, p. 49).
3.2.2. White South African mothers
The white Afrikaner population in South Africa is mainly descendants from the Dutch and British as a result of colonialisation. The role that was allocated to the Afrikaner women in the late 19th and early 20th century was that of ‘Volksmoeder’ and became an important component in the propaganda arsenal of Afrikaner nationalism. The Volksmoeder was the cornerstone of the household and was identified with characteristics like ‘kindness, gentleness, care, frugality, discipline and conformity’ (Brink, 1990). According the Volksmoeder ideology, women were perceived to place family and social needs before their individual preferences, and their status was based on their success as housewives. Like the Western intensive motherhood ideology, this ideology expects very separate spheres of activity for men and women. The ideology was based on essentially white middle-class values and imposed on the working class. Thus, both middle-class and working-class mothers identified with the ideals of the Volksmoeder, but working-class mothers could not often not meet the Volksmoeder ideals (Brink, 1990).
In 2005, a relatively high gender gap of labour force participation (16%) was reported within the white South African population (Van der Westhuizen et al., 2007, p. 10). This means that significantly more men are participating in the labour force than women. In addition, the increase of labour force participation in white women (almost 9%) was significantly lower than that in African and Asian women (Van der Westhuizen et al., 2007, p. 10). These numbers are in accordance with the separate spheres linked to the intensive motherhood ideology. Moreover, the intensive motherhood ideology is historically related to the nuclear family. Within the white South African population, the nuclear family structure was by far the most common family form in 2005 (63%) and
much more prevalent than extended families (12%). In addition, a relatively high percentage of the white households were single-parent households (17%) (Amoateng et al., 2007, p. 49).
3.2.3. Asian South African mothers
The Asian population in South Africa mainly descended from Indians who arrived since 1860 to work on the country’s sugarcane plantations and is strongly connected to the collective Indian culture and religion. Within the Indian culture, three types of family structures exist. First, the traditional joint family includes family members of different generations. Second, the extended family refers to a family in which married sons and brothers live separately but continue to have joint property and share income. Third, the nuclear family consists of the male head of the household, his wife, and unmarried children (Bahadur & Nisha, 2008, p. 74). In South Africa, almost 60% of the Asian population lived within nuclear families in 2001. Compared to 1996, this number decreased by approximately 3% (Amoateng et al., 2007, p. 49).
Despite the normality of extended families, traditional Indian family roles are well defined. The role of the Indian women has many similarities with the Western ideal of intensive motherhood. Traditionally, Indian women are seen as responsible for the domestic work and childcare while men work outside the home (Srinivas, 2006, p. 198). The traditional Indian ideal of womanhood is linked to unselfish motherhood, and it is argued that Indian women are raised to fulfil this prescribed role of mother (Sujatha & Gokilavani, 2011, p.147). In 2005, the gender labour force gap within the Asian population was more than 25%, the highest of all groups (Van der Westhuizen et al., 2007, p. 10). Compared to Asian men, South African Asian women are less present within the public domain. However, the modernisation and urbanisation of the 21th century have exposed Indian women to educational resources, and more Indian women are entering the workplace (Nangalia & Billet, 2009). The female labour force participation rate increased by 14.5% between 1995–2005 (Van der Westhuizen et al., 2007, p. 10). These changes are related to alterations in family structures and new cultural and subjective meanings of gender roles (Srinivas, 2006, p. 199). Some researchers argued that nuclear family structures are today more common within the Indian population but that ‘togetherness’ is still an important aspect (Chadha, 1999, p. 2). In South Africa, most of the Asian population lives in a nuclear family, but the number of extended Indian families slightly grew between 1996–2001 from approximately 28% to more than 30%. Of the Asian group, only 8% of the households were single (Amoateng et al., 2007, p. 49).
3.2.4. Coloured South African mothers
The coloured population in South Africa is a very diverse group because the group includes people who are descendants of different racial groups. As a result, the dominant motherhood ideology among this group is difficult to estimate. However, according to the 2005 data, its female labour force participation is relatively high, with the lowest participation gap of all groups (6%). From these numbers, motherhood seems to be
integrated with employment. The increase in coloured female labour force participation between 1995–2005 was relatively small (8.8%) (Van der Westhuizen et al., 2007, p. 10). The family structures of the coloured population are also diverse. In 2001, slightly more than 50% of coloured families were nuclear families, and 35% were extended families. Among all racial groups, the coloured population had the lowest number of single households (Amoateng et al., 2007, p. 49).
3.3. Research site
This research focuses on the city of Cape Town, the second largest city of South Africa and the capital of the Western Cape province. Cape Town is a relatively wealthy city, although most inhabitants live in poverty. Wealth is concentrated in the city’s northern and southern suburbs, while the poorest people live in townships situated on the periphery of the Cape Flats (De Swardt, Puoane, Chopra, & Du Toit, 2005, p. 101). The economic differences between racial groups are very broad (De Swardt et al., p. 102). A relatively low percentage of the population in Cape Town is identified as black. Although black inhabitants are still the largest racial group in the city (42.6%), they have a significantly lower representation than the national average of almost 80%. By contrast, the coloured population in Cape Town (39.9%) more than four times the national average and almost as large as the city’s black population. In addition, 16.5% of the city’s population are white and 1.9% are considered Asian (Small, 2017, p. 2).
3.4. Contemporary South African motherhood
The limited literature on South African motherhood stresses that historically South African women are deeply divided. According to Magwaza (2003), the apartheid era was responsible for the lack of diffusion of mothering ideologies in South Africa. Since democracy was established in 1994, the mothering patterns have been changing because people of different races and related motherhood ideologies mingle more (Magwaza, 2003, p. 11). Insights into the contemporary context and conceptualisation of South African motherhood have not be explored in the literature (Moore, 2013, p. 156). Despite far-reaching social changes in South African society, the literature assumes that the constructions of South African motherhood have not changed. In particular, women in urban areas of South Africa are exposed to a whole new range of responsibilities and social identities, which affect the understanding of ‘good’ mothering. However, only limited research has focused on the contemporary conceptualisation of motherhood (Walker, 1995).
In South Africa, motherhood is relatively underresearched. The existing studies on South African motherhood have focused mainly on disadvantaged mothers in difficult situations, including those who are single, teenagers, depressed, or incarcerated (Frizelle & Kell, 2010, p. 26-27). To date, the literature on South African motherhood lacks examination of “contemporary regular mothering experiences and practices in its multi-layered and complex nature” (Kruger, 2006, p. 194).
Moore (2013) conducted research on the changing character of motherhood in South Africa by looking at three generations of mothers. Moore’s (2013) results showed that mothers from the youngest generation have manoeuvred from a ‘good provider and caring role’ to a stronger focus on achieving personal ambitions (p. 151). Moore’s research (2013) suggests that motherhood is subject to change and that the intensive motherhood ideology no longer explains South African mothers’ decisions. Christopher (2012) referred to this as a shift from the intensive motherhood ideology to the extensive motherhood ideology. In contrast to intensive motherhood, extensive mothering reflects the idea that a mother’s own needs are as important as the needs of her children. ‘Extensive mothers’ work outside the home mainly of their own interest. Although intensive and extensive motherhood are different, the ideologies are compatible because mothers can retain authority and ultimate responsibility for their children while delegating the tasks of childcare to others.
Macleod (2006) also highlighted that mothering exists in great variabilities across and within historical periods and societies and that it is responsive to its environment. Therefore, mothering is not stable but changeable (p. 128). To understand the conceptualisation of South African motherhood, (historically) social, economic, and cultural differences of social groups must be taken into account (Iwelunmor, Nompumelelo, & Collins, 2010, p. 1393). By including mothers with different socioeconomic and cultural backgrounds, this research aimed to examine how various factors such as class, race, and family structure influence the contemporary conceptualisation of ‘good’ motherhood.
4. Methodology
To understand the contemporary motherhood rationalities of South African mothers, this research used a mixed-methods approach. This approach was used to capture the conscious and subconscious understandings of participants concerning ‘good’ motherhood. Within cognitive science, the stratified nature of cognition is widely shared idea, and cultural cognitivists show that “most of our cognitions occur on the level below of conscious awareness” (Vaisey, 2009, p. 1681). Thus, it was important to gain insights into subconscious rationalities.
For a long time, interviews were thought to primarily engage with discursive consciousness while providing limited insights into deeply internalised moral schemas (Vaisey, 2009, p. 1688). Forced-choice research techniques were considered more useful because “choosing from a fixed list of responses is akin to solving a practical problem […], fixed-response survey questions may draw disproportionately on practical consciousness, which has to make (as opposed to discuss) many such decisions each day” (Vaisey, 2009, p. 1688-1689). Thus, forced-choice approaches were thought to provide insights into practical choices because they trigger internalised cultural scripts.
Today, many cultural cognitivists believe that the conscious and subconscious cognitive pathways are related (Pugh, 2013; Vaisey, 2009). “Given that [both pathways] reside in the same person, who has the same experiences, upbringing and resources, it is not inconceivable that the two levels might be linked in some way” (Pugh, 2013, p. 46). That insight suggests that interviews can play a more important role in understanding unconscious rationalities than originally assumed. Pugh (2013) argued that “interviews are particularly helpful to excavate and interpret emotions, which serve to animate, situate, and connect the levels of consciousness” (p. 44). Thus, in-depth interviews (subjected to interpretive analysis) can capture different levels of information and be considered a useful tool to obtain access to ‘practical consciousness’.
This research used a mixed-methods approach to get a multi-dimension understanding of gender moral rationalities. To gain insights into the conceptualisation of ‘good’ motherhood of South African mothers, semi-structured interviews combined with forced-choice assessments were conducted among a diverse group of South African mothers. However, the results of this research were primarily based on qualitative data from the interviews and discussions that were linked to the ordering of moral statements about motherhood because the quantitative data turned out not to measure motherhood rationalities adequately. On this is reflected in section 4.6.
4.1. Participant sampling & recruitment
To measure the gendered moral rationalities of South African mothers, a diverse group of participants were selected. The participant selection criteria were as follows: women who were born in South Africa, currently live in Cape Town, and have at least one child under the age of 19. To ensure diversity within the sampling group, they were selected on two more characteristics: (1) working/non-working and (2) economic disadvantaged/economic advantaged. The distinction between working and
non-working mothers was made on the basis of self-identification. All mothers who identified themselves as ‘working’, worked substantial part-time or full-time paid hours. However, mothers who identified themselves as stay-at-home mothers frequently had part-time (volunteer) jobs (which they often did from home) with short hours.
Furthermore, to measure class, the selection instrument ‘type of work’ could not be used in this research since stay-at-home mothers were also included. ‘Education level’ also did not always match with lifestyle (which often depends on the economic situation). Some mothers did not have much education but were considered middle-class because they were, for example, married to a rich man and lived in fancy houses in relative financial wealth. Others had quite a high level of education but due to circumstances had a low standard of living. ‘Place of living’ gave the best insight into which class group a mother belongs. Mothers from the townships were categorised as economically disadvantaged while mothers who lived in the suburbs of Cape Town were considered middle class/economically advantaged.
In total, 26 women participated in both the qualitative and the quantitative parts of this research. These participants were recruited through purposive convenience sampling. My personal and professional network in Cape Town was used to recruit some key informants. In some cases, the key informants were themselves mothers, and when they met the selection criteria, they were asked for their consent to interview. However, in all cases, the key informants were asked to provide contact information of (other) mothers who met the selection criteria. After the key contacts obtained permission from their contacts to share contact information with me, the appointed mothers were approached through email, phone, or WhatsApp. During this first contact, the purpose of the research was explained, and the mothers had the opportunity to ask questions. In addition, the mothers were checked to ensure they met the selection criteria. Some mothers refrained from participating or were ultimately found not to meet the selection requirements. For the mothers who met the selection criteria and were willing to participate in the study, appointments were scheduled at times that suited them. After the interviews, the participants were asked whether they knew other mothers that possibly could be interviewed. Thus, a snowball sampling was used to make contact with acquaintances of the existing participants and recruit more participants (Bryman, 2012, p. 511).
4.2. Participant group
The participant group differed in age, employment status, socioeconomic status, education level, family structure, number and age of the children, race, and religion. The group consists of 26 South African mothers aged between 20 and 57 years, with an average age of 36 years.
Slightly more than half of the participants (57.7%) were working mothers and most of them were working full-time (12). One of the mothers was currently studying, and 42.3% of the mothers identified themselves as stay-at-home mothers. Given the high level of unemployment in South Africa, it is important that this research included
unemployed mothers. Of the stay-at-home mothers, 4 were looking for a paid job and the rest (7) voluntarily stayed at home.
The participants lived in many different neighbourhoods within Cape Town. Most participants (14) lived in the relatively prosperous suburbs: Claremont (2), Fish Hoek (2), Gordons Bay, Mowbray, Newlands, Rondebosch (east), Salt River, Sea Point, and Table View (4). The rest of the mothers (12) had a home in one of the following poor townships: Blue Downs/Eerste Rivier, Gugulethu, Kayelitsha (3), Lavender Hill (3), Mfuleni, Nyanga, and Pinati (2).
In addition, the education level of the participants differed from some years of primary school up to an almost finished PhD. Of all mothers, six had not finished their primary school education. Five of the mothers had achieved their matriculation, with which they successfully completed primary school. In addition, nine mothers studied after primary school and achieved a certificate, diploma, or degree. Five of the mothers studied further and finished an honours or Master’s degree. Some mothers are currently completing a PhD. For more information on the South African schooling system see Appendix 1.
Furthermore, the participants live in different family structures. The most common family structure was the nuclear family (61.5%). Traditionally, this family consists of a mother, father and child(ren) that are biologically connected. However, this sample had more variation with the nuclear family structure. In some cases, family members were not all biologically connected. In other cases, the nuclear family was incomplete because some children lived with other family members in another household. In addition, a significant proportion of mothers did not have a partner (34.5%). These mothers lived in single households (7.6%), but most of them were a part of an extended family (26.9%). In contrast to the nuclear family structure, an extended family consist of more than two generations in one household. In total, 30.7% of all family structures in the sample were extended families. Besides single mothers, one couple lived in an extended family structure. Overall, the sample included mothers with a total of 45 children. Most mothers had two children, and 10 mothers had only one child. One mother had three children, and the mother with the most children had four children. On average, the number of children per mother was 1.7. Slightly more than half of all children were boys (24), and the rest (21) were girls. All children were between the ages of 3 months and 40 years, with an average age of 10.8 years. However, the focus in this research was on mothers with children up to and including 18 years. Thus, when including only children between 0 and 18 years, the average age was slightly lower (8.1 years).
The vast majority of the participants racially identified themselves as black or coloured/mixed (both 38.5% of the mothers). Four participants identified themselves as white, and two mothers considered themselves as Cape Malay. In addition, 12 mothers identified themselves as Christians. Five mothers considered themselves as Muslim, and four mothers argued they feel more spiritual than religious. The minority group comprised adherents of Judaism (1), the New Apostolic Church (1), and the Revelation
Church (1). Only two mothers claimed that they have no religious or spiritual connection.
Within the participant group, some social demographic factors seem to be correlated. Relationships were found between education level, racial background, and family structure. Educational level seems to be linked to race. Only mothers who identified themselves as coloured or black did not finish primary school. However, mothers who identify as coloured often achieved a certificate, diploma, or degree. Additionally, mothers who live in a nuclear family often had substantially higher level of education. Thus, almost all mothers that consider themselves as white lived in a nuclear family structure, while mothers who identify as black were often single or lived in extended family settings. Furthermore, family structures were partly connected to class. The vast majority of nuclear families lived in the suburbs, while extended families were overrepresented in the poor townships. This is not surprising because neighbourhood and education seemed to also be strongly connected. Therefore, differences in class are also racially patterned.
In addition, being religious seemed to be linked with the economic disadvantaged class. Non-religious mothers or spiritual mothers were all economic advantaged mothers, while all economically disadvantaged mothers reported that they are religious (all Christians). Furthermore, the sample seemed to show a correlation between Christianity, being economically disadvantaged, and racial background. That is, all economic disadvantaged mothers in the sample were Christians who identify as racially black or coloured. Additionally, all mothers that racially identified as white were stay-at-home middle-class mothers. Appendix 2 lists more detailed participant information.
4.3. Data collection & operationalisation
The meetings with the participants consisted of two parts: a qualitative part, followed by a quantitative part. For most participants, these two parts were conducted after each other during the same meeting. In some exceptional cases, the meeting was split up into two due to time pressure. The participants agreed to a location where the meeting would take place, and depending on their preference, the meetings were held in my student room, restaurants, cafes, fast food restaurants, parks, and participants’ homes or work places. The interviews and discussions lasted between 1.5 and 2.5 hours. The language used during the meeting was English, and although this language was not the mother tongue for me and some participants, everyone was able to express themselves properly, partly because English is one of the official South African languages.
During the first part of the meeting, the participants were interviewed. Semi-structured interviews were used to give voice to South African mothers so that the underlying moral mechanisms of motherhood rationalities could be better understood. Semi-structured interviews were selected because they enable for both induction and deduction: besides the established discussion topics, there was also space for the participants to bring in their own themes.
The topic list of the interviews drew upon the existing literature on motherhood norms. Because only limited research exists on contemporary South African motherhood (Frizelle & Kell, 2010, p. 26-27), the topics also derived from Western literature on motherhood. As described in the theoretical framework (chapter 2), the conceptualisation of ‘good’ motherhood is related to mother’s understandings of what children need and how these needs should be met. In addition, the ‘right’ division of labour refers to mothers’ understandings of how these children’s needs should be balanced with their own needs.
To gain insights into the conceptualisation of ‘good’ motherhood and the division of labour, this research focused on mothers’ understandings about the following topics: children’s needs and the organisation of childcare, the role of employment, family structures, the provision of income and the importance of financial independent, the role of a partner or social network, outsourcing of childcare, family and religious traditions, and gender norms and visions on motherhood policy and regulations. Thus, to understand the moral rationalities concerning motherhood and the (preferred) relationship between paid work and childcare, the interviews focused on motherhood practices and ideals. Appendix 3 presents more information about the interview topic list.
In addition, to gain insights into the mothers’ unconscious part of valuation, the Q-method of Stephenson (1935) was used. The Q-Q-method enables the transformation of subjective values and moral perspectives into operant factors. The participants were asked to organise 25 statements with moral considerations related to family life and employment during the second part of the meeting. These normative statements contained the same issues as the topic list for the interviews and linked the different issues to ‘good’ motherhood. The used statements are listed in Appendix 4. The participants ordered the statements on a normal distribution diagram, with ‘totally disagree’ on one side and ‘totally agree’ on the other side. The statement on the ‘totally disagree’ side was labelled with –4. Most of the statements could be put in the middle of the normal distribution, these neutral statements were labelled with 0. The statement that was put at ‘totally agree’ was labelled with 4. All the other statements got a label between −4 and 4. The statement ordering of the participants were photographed, and the labelled data was entered in SPSS. After the ordering, the statements were discussed one by one and, if possible, also linked to earlier statements made during the interviews. Thus, the second part produced not only quantitative data but also qualitative data.
4.4. Analysis of the data
All qualitative data of the interviews and statement discussions were recorded and transcribed. The research software ATLAS.ti was used to facilitate the coding and analysis process. First, the qualitative scripts were coded using conventional content analysis. This coding method derives categories inductively from the data and is mainly useful when there is limited research on a topic (Hsieh & Shannon, 2005, p. 1279). In this case, using conventional content analysis avoided coding based primarily on
