**Life Course Research and Social Policies 8**

## Laura Bernardi Dimitri Mortelmans *Editors*

# Lone Parenthood in the Life Course

## **Life Course Research and Social Policies**

Volume 8

**Series editors** Laura Bernardi Dario Spini Jean-Michel Bonvin Life course research has been developing quickly these last decades for good reasons. Life course approaches focus on essential questions about individuals' trajectories, longitudinal analyses, cross-fertilization across disciplines like lifespan psychology, developmental social psychology, sociology of the life course, social demography, socio-economics, social history. Life course is also at the crossroads of several fields of specialization like family and social relationships, migration, education, professional training and employment, and health. This Series invites academic scholars to present theoretical, methodological, and empirical advances in the analysis of the life course, and to elaborate on possible implications for society and social policies applications.

More information about this series at http://www.springer.com/series/10158

Laura Bernardi • Dimitri Mortelmans Editors

# Lone Parenthood in the Life Course

*Editors* Laura Bernardi National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life course perspectives (NCCR Lives), University of Lausanne Lausanne, Switzerland

Dimitri Mortelmans CLLS, Faculty of Social Sciences University of Antwerp Antwerp, Belgium

ISSN 2211-7776 ISSN 2211-7784 (electronic) Life Course Research and Social Policies ISBN 978-3-319-63293-3 ISBN 978-3-319-63295-7 (eBook) DOI 10.1007/978-3-319-63295-7

Library of Congress Control Number: 2017952361

© The Editor(s) (if applicable) and The Author(s) 2018. This book is an open access publication. **Open Access** This book is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this book are included in the book's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the book's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.

The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG

The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

## **Contents**



## **List of Figures**



#### List of Figures


## **List of Tables**



List of Tables


## **Chapter 1 Changing Lone Parents, Changing Life Courses**

**Laura Bernardi, Dimitri Mortelmans, and Ornella Larenza**

#### **Changing Pathways of Lone Parents in Europe**

The socio-demographic profile of lone parents has changed in the last decades. Being mostly widowed men and women or young single mothers until the 1970s, lone parents are nowadays mostly divorced and separated parents, even though they are still by and large mothers rather than fathers. As a consequence, the experience of lone parenthood has also dramatically changed. Less objects of pity or stigmatized with shame, lone parents and their children are more than ever bound by legal arrangements to the other parent and are caught in more dynamic family trajectories.

There are at least two remarkable changes that certainly need to be addressed by research on lone parenthood: its boundaries and its diversity. Both aspects are connected and have potential implications for lone parents and their children. First, the diversity and complexity of legal and residential arrangements of parents and children make it difficult to establish the borders between a full-time and a part-time one-parent household. When child custody or parental authority are shared, can we still talk about lone parents? Children circulate more and more between two or more parental households after separation, and more than one parent may be financially and legally responsible for them. One direct consequence of such changes in the

L. Bernardi (\*) O. Larenza

The original version of this chapter was revised.

An erratum to this chapter can be found at https://doi.org/10.1007/978-3-319-63295-7\_16

National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life course perspectives (NCCR Lives), University of Lausanne, Lausanne, Switzerland e-mail: Laura.Bernardi@unil.ch

D. Mortelmans CLLS, Faculty of Social Sciences, University of Antwerp, Antwerp, Belgium

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_1

phenomenon of lone parenthood is that it is not straightforward to establish even basic descriptive statistics on lone parents across countries and datasets.

Second, the growing likelihood of re-partnering changed lone parenthood into a more temporary phase in the life course. Despite differences in the duration of lone parenthood episodes depending on the gender, the number and age of the children, and the educational and migration background of the lone parent, lone parenthood durations are shorter than in the past. Yet, re-partnering does not always mean the creation of a new residential unit with cohabiting partners; living apart together with a new partner is not rare among separated and divorced parents. In case the nonresident new partner takes up part of the financial and parenting responsibilities, can we still talk about lone parenthood? Boundaries of the definition and complexity of the relationships concerning lone parenthood are just two aspects that exemplify the challenges facing research on lone parenthood in the XXI century (see the Chap. 2 by Letablier and Wall for a systematic discussion of definitions).

This introduction gives first an overview of the recent trends in lone parenthood across Europe filling a gap in the scientific empirical literature on lone parenthood, which is rarely comparative and rather dated by now (with the exception of the recent report on lone parents in the UK by Berrington 2014). Second, it gives an overview on the literature on lone parents in relation to other life course domains like employment, health, poverty, and migration. We also touch on parenting and children's outcomes. We conclude with a brief discussion on the universalistic and targeted welfare approaches to meet those lone parents in need of support. We hold that the current volume represents a first step to relaunch research on lone parenthood in the XXI century through a life course perspective. This is much needed updated knowledge and reflection on a changing phenomenon given that an ever-greater number of children spend at least part of their childhoods in one-parent households, their social background and needs are more heterogeneous than in the past, their relation with parents and grandparents are increasingly complex, and last but not least, the institutional context in which their family lives impacts children's life chances significantly.

With the spread of union disruptions, an ever-greater number of children grow up during at least a part of their childhood in a one-parent household because many of them live in increasingly complex families, because their social background and their needs are more and more heterogeneous, and because the institutional context in which their parents live has important consequences for how lone parents and their children fare in comparison with other families.

#### **Prevalence of Lone Parents in Europe**

The phenomenon of lone parents as a social group that deserves special attention by policies arose during the nineties when one-parent households became statistically visible (Bradshaw et al. 2000; Kennedy et al. 1996). Several studies have made calculations of the prevalence of one-parent households throughout Europe and other OECD countries. Unfortunately, most of these rates differ a lot according to the source being used. Most international comparative surveys have been used to look at lone parenthood: ECHP (Chambaz 2001), PISA (Chapple 2009), LIS (OECD 2015), and EU-SILC (Iacovou and Skew 2010). Some rates are calculated among the percentage of families with children (OECD 2011), often because the survey on which it is based contains only families with children (Chapple 2009). Also, definitions are often not exactly the same. Sometimes children are counted until the age of 15 (Chapple 2009), 18 (Iacovou and Skew 2010), or 25 years (Chambaz 2001). Also, the inclusion of so-called 'included' lone-parent families (those sharing an accommodation with another household) might lead to considerable differences in rates (Chambaz 2001).

All this diversity in previous studies makes it difficult to make comparisons with previous results. In this introduction, we use the Harmonized Histories1 with comparable data on fertility and marital histories from 18. In our analyses, we define lone parents as single living adults in the age range of 15–55 with children aged 18 or younger present in the household.2

Lone parents take on an increasing share of all households throughout the past five decades (Table 1.1). In all countries, we see an increase in the prevalence of lone parenthood even though the cross-country variation is huge. As was shown with other data (Iacovou and Skew 2010; OECD 2011), the USA, the UK, and Russia end up in the top. Sweden has been found to be a high-prevalence country as well but shows only an average rate in our analyses. The low-prevalence countries are southern European countries and Poland, Romania, and Georgia.

In Table 1.2, we look in more detail at the first spell of lone parenthood in the life course. The mean age at the first episode of lone parenthood circles around thirty in all countries. Lower mean ages are found in the USA but not in the UK. Given the high proportion of teen pregnancies, we would have expected this mean age to be lower in the UK. Probably this is due to the omission of 'included' lone parents, as most of the teen mothers continue living with their parents and remain unobserved in these analyses. Also, the length of lone parenthood differs considerably across countries, with Switzerland, Georgia, Lithuania, and Russia showing the longest spells of lone parenthood. This may mean that chances of re-partnering for lone parents depend on the local context or an indication that divorce happens at different stages in the life courses (when divorces occur at later ages, children are rightcensored quicker out of the household). A more detailed analysis of lone parenthood durations is shown in Table 1.3, where we consider changes in durations across cohorts by country. Re-partnering chances for lone parents seem to have increased over birth cohorts. Most countries experience a jump in re-partnering chances in

<sup>1</sup>The Harmonized Histories data file was created by the Non-Marital Childbearing Network (http:// www.nonmarital.org) (See: Perelli-Harris et al. 2010). It harmonizes childbearing and marital histories from 14 countries in the Generations and Gender Programme (GGP) with data from Spain (Spanish Fertility Survey), the United Kingdom (British Household Panel Study), and the United States (National Survey for Family Growth). Thank you to everyone who helped collect, clean, and harmonize the Harmonized Histories data, especially Karolin Kubisch at MPIDR.

<sup>2</sup>The authors want to thank David Deconinck for his help with the analyses. Emanuela Struffolino has done all calculation in this chapter for Switzerland. We also thank her for this extensive work.


**Table 1.1** Prevalence of lone parenthood in Europe and the USA in % of all households in the country (age group 15–55, period 1960–2010)

Source: Harmonized Histories, v12.10.2015 (Authors' calculation)

either the 1951–1960 cohort or the 1961–1970 cohort. In older cohorts, the average length of lone parenthood approximates 8–10 years, whereas the younger cohorts clearly move towards 4 or even 2 years on average. This is a sign that lone parenthood status is changing in nature. We will elaborate on that in the next paragraph. Table 1.2 also shows some variation according to educational level. There is, however, no clear pattern to be discerned across countries. In some countries, most highly educated clearly have better chances of leaving lone parenthood status, while in other countries those with an intermediary level of education are better off. According to gender, men have much shorter spells of lone parenthood compared to women. Probably, this is a sign of differences in custody arrangements (being more favourable for men in terms of chances on the partner market) or a reflection of the general higher chances of men on the partner market.

The 2011 OECD study revealed that children in lone-parent families are becoming older, while the size of the lone-parent families is shrinking (OECD 2011). We find few country differences in the size of lone-parent families (Table 1.4). Only the UK and the USA have larger lone-parent families, more or less attributable to educational level. The most highly educated lone-parent families are larger than the least educated ones. There is little or no difference in size between a mother-headed and a father-headed lone-parent family.


**Table 1.2** Occurrences of lone parenthood and length of the first occurrence in Europe, by education and sex (age group 15–55; cohorts 1921–1990)

a Respondents in the Harmonized Histories with at least 1 occurrence of lone parenthood between 15 and 55 years

#### **Life Course Trajectories of Lone Parents**

In the previous paragraph, we demonstrated that the average length of the first spell of lone parenthood differs across countries. In this paragraph, we further concentrate on lone parenthood in a life course perspective. Seldom, the dynamic nature of lone parenthood has been shown in cross-country overviews. The prevalence of lone parents in comparative studies usually suggests an instantaneous view on the share of lone parents in one single country. Life course analyses reveal that over time, lone parenthood is a transitionary state, with adults being active on the partner market and successfully engaging in new relationships.

In Table 1.5, we provide more insight in the household composition of lone parents, 1 year before entering lone parenthood and 1 year after exiting this state. Not surprisingly, most lone parents were either married or cohabiting before a relational


**Table 1.3** Length (in years) of first spell of lone parenthood, by birth cohorts (age group 15–55)

Source: Harmonized Histories, v12.10.2015. The data on the two youngest groups of cohorts may be resulting from important right censoring

breakup that left them alone with the children. Countries showing higher rates of cohabitation like Norway and Sweden also generate more lone parents from that state. Only Spain seems to be a peculiar exception to this rule: The vast majority of lone parents were cohabiting before lone parenthood, while Spain is known for very high marriage rates. A possible explanation of this outlier might be some kind of stigma that is associated with cohabitation preventing lone parents from being 'included' in the parental home after a breakup. Chambaz (2001) showed that Spain has higher rates of 'included' lone parenthood. Because these lone parents are masked in our analyses, we might find a statistical artefact here when stigma indeed prevents adult children returning home after a breakup. The first and third columns in Table 1.5 show the prevalence of married or cohabiting partners without children turning into lone parents 1 year later. These refer to a relational breakup of pregnant women which may give an indication of unwanted pregnancies. These rates are typically higher among cohabiters than in marriages. The last column in the table shows adults who become parents outside a relationship. The data are unclear about the exact singleness status 1 year before lone parenthood. We do not observe marriage or cohabitation in the data, but they might represent deliberately single parents or living-apart-together (LAT) relationships. We do not possess any information to make the distinction between these two.


**Table 1.4** Number of children in lone-parent families in Europe, by education and sex (age group 15–55)

a Respondents in the Harmonized Histories with at least 1 occurrence of lone parenthood between 15 and 55 years

Leaving the lone parenthood status goes in two pathways. First, the lone parent re-partners by entering marriage or cohabitation. Again, Sweden, Norway, and the USA show much higher rates of cohabitation than marriage. Overall, in most countries, new relationships are more often cohabitations than marriages. Only in Hungary do we see a very high remarriage rate among lone parents. The second route out of lone parenthood occurs when the parent no longer has children in the household. This could indicate a change in the custody arrangement, but it might as well be an empty nest. Again, the data do not allow us to make this distinction. As Table 1.4 has shown fairly high average ages of lone parents, we assume that most lone parents exit their lone parenthood status when the last child becomes independent.

The last analysis (Table 1.6) further elaborates on lone parents' chances to re-partner. Using Kaplan-Meier estimates, we describe the patterns and tempo of re-partnering among lone parents. This analysis shows that re-partnering occurs at a


**Table 1.5** Status of lone parents 1 year before entry and 1 year after exit (age group 15–55)


**Table 1.6** Timing of re-partnering of lone parents, by country (age group 15–55)

much slower speed in Eastern European countries like Bulgaria, Georgia, Lithuania, and Russia. Other Eastern European countries like Estonia and Romania show particularly quick re-partnering routes. Also, Germany and the UK have partner markets that make it easier for lone parents to re-enter a new cohabitation or marriage. For a majority of lone parents throughout Europe, living alone with children is a state that takes at most 5 years. In almost all countries included in our analyses, we see that half of the population of lone parents re-partners within 5 years, and in a majority of countries, almost three-quarters of lone parents exit the status within 10 years. Of course, the prevalence of empty nests in this 5- to 10-year period increases dramatically, as not only lone parents but also their children age (and partner) during this period.

#### **Challenging Life Domains for Lone Parents: Poverty, Work, and Health**

Even though the lone parenthood state is often a transitory state in the life course, it interferes severely in many life domains. Many studies have pictured the short- and long-term consequences of becoming a lone parent. In this section, we give an overview of the field by focusing on lone parents' work trajectories, health outcomes, poverty risks, and the implication of migration for lone parenthood. This overview only targets the lone parents themselves. In the next paragraph, we expand our overview to the children of lone parents because being a lone parent automatically involves dependent children.

#### *Work Trajectories*

Work, care, and income are a triangle of intertwined influences and dependencies that are difficult to see completely independent from each other. In this and the next two paragraphs, we will summarize the main findings of this complex equilibrium. We will do so from an individual perspective since the institutional influence on work and care is discussed in paragraph 7.

As lone parents are predominantly women, the research on employment patterns of lone parents focuses on female labour participation. Men, whether single, married, or lone parent, show consistently successful and stable employment histories and are therefore left out of most studies. In general, there is an increase in the female labour market participation across OECD countries. In most countries, the employment rate surpasses 60% (OECD 2011). Scandinavian countries are well above this average, while the UK is characterized by a rather low employment rate among lone parents (especially lone mothers) (Millar and Rowlingson 2001). The UK shows a double outlier position, as the country has not only lower rates compared to other countries but also shows an employment gap within the country compared to employment rates of married women (Gregg and Harkness 2003). A caveat on these rates, however, is that even though single mothers in many countries have higher employment rates, they do more part-time work than mothers in partnership or childless single women (Ruggeri and Bird 2014).

The presence of lone mothers on the labour market has been shown to be highly selective. First, lone mothers are more likely to be less educated leading to trajectories in lower-skilled and worse-paid jobs (Stewart 2009; Zagel 2014; Zhan and Pandey 2004). Being stuck in the lower-skilled and lower-paid jobs also implies an increase in what Bauman (2002) calls the hidden costs of employment: increased financial costs for childcare, higher transportation costs, and a decrease in available hours for direct household needs. Second, the household composition itself plays a role. The higher the number of children and the younger these children are, the less likely a lone mother will be to work. The effect is found consistently in the literature and is plausible. The more children there are and the younger the children, the more the time budget of the lone mother is constrained (Drobnic 2000; Hancioglu and Hartmann 2012). Both the costs for young children not going to kindergarten or school and the costs for an increasing number of children in the household are a significant burden to the labour supply of lone mothers. Third, life course researchers also point to the moment in the life course where lone parenthood occurs. Both finding or keeping one's job is more difficult when experiencing lone motherhood at a young age than when experiencing lone motherhood at middle age (Chevalier et al. 2003). Fourth, social benefits like single-parent allowances also turn out to have a negative effect on the labour supply of lone mothers. As the financial burden after the breakup is softened, women are less encouraged to increase their labour supply (van Damme et al. 2009).

Following the work of Hakim (2000) on preferences of women for work or care, several authors have looked into the issue of whether it is the preferences or attitudes of women that keep them off the labour market. These authors call attention to the cultural dimension of labour supply instead of focusing only on the structural and institutional factors (Gingerbread 2012). One prominent response to Hakim's criticized typology is worked out by Bell et al. (2005). They position labour market decision on two axes: (1) a work orientation axis and (2) a parental care orientation axis. This double approach shows that women can have both high aspirations on the labour market and high aspirations to care, which was impossible in Hakim's typology. They also construct the typology with a dynamic perspective, as aspirations can change over the life course. A more recent study by Boeckmann et al. (2014) confirms the importance of looking at cultural norms regarding maternal employment in addition to individual and system characteristics.

#### *Poverty*

Probably one of the firmest associations of lone parenthood in both policy circles and in academia is that with poverty. There is a huge amount of evidence that the state of lone parenthood is associated with poverty. One of the most comprehensive overviews by the OECD (2011) shows poverty rates among lone parents ranging from less than 10% to over 40%. The lowest poverty rates are found in the Scandinavian countries (Sweden and Denmark), while the highest numbers come from the US, Australia, and Canada. In Europe, high poverty figures are also found in Spain, Germany, Estonia, the UK, and Ireland. The main foci in the research on how to tackle these extraordinary poverty figures among lone parents are either on income taxes (Brady and Burroway 2010; Heuveline et al. 2003) or on transfers and family policies (Maldonado and Nieuwenhuis 2015; Van Lancker et al. 2015). From both perspectives, results suggest that universal transfers are more effective for lone parents than measures targeting them as a specific group (Brady and Burroway 2010; Chzhen and Bradshaw 2012; Maldonado and Nieuwenhuis 2015).

Many factors that explain this increase poverty risk have already been documented. On the one hand, factors associated with work potential (see section "Parenting") like low-paid jobs and part-time work contribute to an increase in poverty risk. On the other, family-related factors like large families or families with young children and low human capital also contribute to the risk of poverty for lone parents. Also, inadequate payment of child support by the ex-spouse deprives lone parents of necessary sources of income (Zhan and Pandey 2004). It shows the close interrelation between labour market participation and poverty across countries (Maldonado and Nieuwenhuis 2015; Ritakallio and Bradshaw 2006). Again, life course researchers have pointed to the dynamic nature of this phenomenon. Placing the poverty risk in a life course framework showed the accumulated disadvantage of more vulnerable social groups (less educated, precariously employed, in bad health) and the reinforcing character of episodes of lone parenthood therein (Tsakloglou and Papadopoulos 2002; Vandecasteele 2010, 2011).

#### *Health Trajectories*

The association between poverty or social inequality and (bad) health conditions has received firm empirical evidence (e.g., Mackenbach et al. 2008). As the association between lone parenthood and poverty is quite firm, it comes as no surprise that health issues are also correlated with the state of lone parenthood. Evidence has been found for both worse general health conditions (Whitehead et al. 2000) and increases in mental health issues (Crosier et al. 2007). The central question in this respect is whether the impoverished background of many lone parents is the main reason for the bad health or the status of lone parent as such is responsible for the worse health conditions (Benzeval 1998). The first link is documented elaborately in the literature on health inequalities and will not be discussed here. Intrinsic reasons for worse health among lone parents are predominantly related to the stress that accompanies the status of lone parent. The combination of work and care for children in a single headed household is much harder, leading to more stress and health issues. In addition, a lack of social support from the (decreased) social network adds to the strain of being lone parent (Cairney et al. 2003).

More recently, much attention went to the buffering influence of welfare states in the association between lone parenthood and health on the one hand and the changes in such association over time on the other hand (and its relatedness to economic crises). In three Swedish studies, the temporal shift in health outcomes was put under scrutiny (Burstrom et al. 1999; Fritzell and Burstrom 2006; Fritzell et al. 2007). The studies showed the emergence of lone parents as a vulnerable social group during the nineties and the first decade of the new century. Economic strain is increasingly associated with the worse health conditions of lone parents, even though the Swedish welfare state intervenes substantially. In follow-up comparisons between Sweden, Britain, and Italy, economic conditions turned out to be less important as a main factor to explain bad health, but the synergy between lone motherhood and non-employment turned out to be a highly important mediation factor (Burstrom et al. 2010; Fritzell et al. 2012; Whitehead et al. 2000). The comparative study revealed influences of both policy regimes and a country's culture and tradition. Lone mothers in this respect are considered a 'litmus' test of the interface between family policy systems and health (Burstrom et al. 2010). When extending the number of countries, regional health inequalities appear throughout Europe with CEE countries reporting the worse health conditions for all (lone and cohabiting) mothers, while the Anglo-Saxon regime negatively influenced self-rated health of lone mothers and Bismarckian welfare regimes showing a more devastating effect on mental health of lone mothers (Van de Velde et al. 2014).

#### *Migration and Lone Parenthood*

While there is growing interest in the union behaviour of immigrants (Huschek et al. 2010; Kalter and Schroedter 2010; de Valk and Milewski 2011), little is known about their family structure. In particular, differences between immigrants and natives in the prevalence and incidence of lone parenthood and the consequences of such differences for immigrants' life courses are little investigated.

This state of the art is possibly due to the fact that lone parents are mostly lone mothers, while migration research for a long time had concentrated on the economic migration of men. Until recently, most migrants were men, migrating alone in the first place and followed by wives and children a few years later when this was possible. The number of lone mothers migrating alone with children was limited. In addition, most migrant groups were in relatively stable unions so that, also after migration, in most Western countries, the likelihood of entering lone motherhood for immigrant women was lower than for the native populations (Landale et al. 2011; Milewski and Kulu 2014). Yet, with the feminisation of migration flows (Lutz 2007) and with the diffusion of divorce and its acceptance in most sending countries as well as among the second generations in the receiving countries (Andersson et al. 2015), studies on immigrant lone parents are much needed. In addition, migrants have specific pathways to lone parenthood. One of them may be related to an unfavourable marriage market for immigrants; the high unemployment rates of immigrants often translate into fewer opportunities for marriage for immigrant women. Under such circumstances, the literature on lone mothers in minority groups suggests that women may then put motherhood before marriage (Edin and Kefalas 2005; Lichter et al. 1992; McLaughlin et al. 1992). Another specific pathway to lone parenthood for immigrants may be related to the process of migration itself: an immigrant woman may enter the destination country while her male partner stays behind, either because he is entrapped in civil wars or because of time-lasting employment obligations in the country of origin (Landale et al. 2011).

Using PISA data in a cross-country comparative perspective, a recent study establishes that the risk of being a lone mother in the country of immigration is positively and significantly correlated with the prevalence of lone mothers in the country of origin but not with that of the country of destination (Dronkers and Kalmijn 2015). The same study concludes that immigrant mothers with a partner who were born in the destination country (in a mixed marriage) and who speak the destination language at home with the children are at higher risk of being lone mothers when their children are 15. Interestingly, the same study finds that while for all children of lone parents, immigrants and natives, there is a disadvantage in school performance with respect to children living in two-parent families, such a gap is smaller in the immigrant population. They interpret all their findings as a proof that immigrant lone mothers are positively selected for being more integrated in the culture of the destination country with respect to partnered immigrant mothers. Those kinds of studies call for future research on the interrelation between lone parenthood and migration.

Yet, research in this field faces challenges related to data availability, in particular when they want to explore pathways in and out of lone parenthood and follow the trajectories of migrant lone parents. The relatively small number of migrants in general panel samples, and their higher probability of dropping out of panel designs, often hinder the examination of the migrant population of lone parents by generation or ethnic group, or by important characteristics like the presence or absence of institutional and social support.

#### **Lone Parents and Their Children**

#### *Parenting*

Two contemporary trends in the employment and family domains combine to impact mothers' care practices and parenting. One trend is the spread of parental employment, especially of mothers' employment, and the corresponding secular decrease in the percentage of nonworking parents (Fox et al. 2013). The second trend is the parallel trend in the banalisation of separation and divorces and the increasing share of children who live, at least for some time, in one-parent households, generally with the mother. As a consequence, lone mothers often have the sole or most of the responsibility for caring and providing for their children, managing the household, and organising childcare during employment activities. How do lone mothers organize care, especially when they work? And how does lone motherhood affect parenting practices?

Lone mothers enjoy less social support than mothers in couples (Mac Lanahan et al. 1981; Amato 1993). The greater care burden among single mothers may also leave less room for the more enjoyable and rewarding aspects of parenting, especially for non-working mothers. Past evidence has suggested that lone mothers' parenting is characterized by less parental engagement (Carlson and Berger 2013) despite the fact that lone mothers spend more solo time with their children than mothers in couple (Kalil et al. 2014). Nevertheless, the effects of lone motherhood on parenting are all but straightforward. Research has highlighted both positive and negative effects of lone motherhood on mother-child interaction. The time squeeze due to mothers' employment and care responsibilities may create feelings of inadequacy and have spillover effects on the relationship and interactions with children (Blair-Loy 2003; Nomaguchi et al. 2005).

At the same time, the consciousness of being able to provide economically for the children and the engagement in activities and relations other than care may have positive spillover effects on the quality of time spent with children (Garey 1999; Latshaw and Hale 2016; Yetis-Bayraktar et al. 2013). In addition, time with children alone can produce very positive feedback on mothers' well-being, providing mothers with a source of meaning related to their role as provider and carer as indispensable (Edin and Kefalas 2005; Villalobos 2014).

A recent study whose aim was to assess the role of partnership status and employment on mothers' well-being and emotions related to parenting, found that, lone mothers report in general less happiness and more sadness, stress, and fatigue in parenting than mothers living in couple. Yet, employment seems to be an important mediator, since employed lone mothers seem to be happier and less stressed than those who are not employed (Meier et al. 2016). Lone mothers seem to suffer from worse subjective health particularly when they have limited labour market prospects, perhaps relying on low-skill part-time employment contracts (Struffolino et al. 2016) or occasional jobs (Campbell et al. 2016). Qualitative research aiming at understanding how lone parents prioritize their responsibilities shows that when employment time conflicts with sole responsibility for parenting (because of lack of childcare), caring obligations take priority. As a consequence of the double responsibility, this burden may result in exhaustion with multiple roles and harsher parenting (Breitkreuz et al. 2010) as well as the inability to supervise children. Parents express concerns about the safety of their children when older children are required to look after younger siblings (Hildebrandt and Kelber 2005). A second consequence in such cases is that the ability to work or maintain employment may be impaired, entailing a vicious circle of disadvantages (Good Gingrich 2010).

#### *Child Outcomes*

Studies have repeatedly shown a clear empirical association between growing up in a one-parent household and poor child outcomes, especially when such a family structure results in a drop in income and in parental involvement (McLanahan and Sandefur 1994). Children of divorce and living in one-parent households have consistently worse educational outcomes and lower levels of well-being compared to children of intact families (Amato and Keith 1991; Amato 2005). In addition, in a life course perspective, there is evidence for the intergenerational transmission of lone parenthood. Children growing up with lone parents have higher chances of experiencing lone parenthood themselves, whether because of transmission of teenage pregnancies or of separation and divorce (MacLanahan and Sandefur 1994).

The observed association between family structure and negative child outcomes has been explained with both causation and selection mechanisms (Amato and James 2010; Thomson and McLanahan 2012). The most common causal explanations given to explain research findings are the higher economic hardship of onehousehold families, the poorer parenting quality of lone mothers, and the more frequent exposure to stress of their children (Amato 2005). Such explanations have in common that it is the fact of living in a one-parent household that produces negative consequences for children in most countries.

A different set of explanations look at parents' characteristics, like their genetic, social, cultural, and economic resources, and posit that these may be the important determinants of children outcomes. Resource-poor individuals may be selected to become lone parents either by not entering couples or being more likely to see their couple break down. For instance, a longitudinal study on single mothers who enter parenthood without a partner showed they suffered from socio-economic disadvantages before becoming mothers (Baranowska-Rataj et al. 2013). In such cases, it would not be the condition of lone parenthood to affect their children, but rather the selection into lone motherhood of disadvantaged women. Parents may transmit disadvantageous traits to their children either through genetic inheritance or inadequate interaction with their children. In this case, the observed association of growing up in a one-parent family and negative child outcomes may be a spurious effect of some inherited parental characteristic influencing both. Selection mechanisms are difficult to test and to rule out without longitudinal or quasi-experiment data and may be involved to a certain extent.

Empirical evidence has so far consistently supported causal explanations indicating that it is growing up with one parent that has put children at higher risk of having cognitive, emotional, and social problems (Amato 2005). Yet, it is an open question for future research whether, as a consequence of the diffusion of one-parent households across social strata and the average shorter duration of lone parenthood episodes (which means shorter exposures to stress, poor parenting and economic hardships for children), the relative influence of causal mechanisms will weaken and selection will become more important (see Chap. 15 by Hannan in this volume).

#### **Welfare States and the Support for Lone Parents**

Two strategies are available to policy makers when it comes to tackling lone parenthood negative outcomes. First, universalistic types of intervention do not target lone-parent households as such, but address those negative outcomes within the general population (e.g., any universal health care system providing services on the basis of a universal right to receive health care). Second, targeted types of intervention tackle lone parent households as their only public (e.g., advances of maintenance payments) in order to target those negative outcomes that are typically associated with lone parenthood. It should be clear, though, that no welfare state entirely fits one of the two models as, in fact, policies are diverse and different strategies can be adopted within different policy fields (e.g., the UK blends a universal health care system and a "New Deal for Lone Parents", including a set of measures intended to increase lone parents' labour market attachment).

Two main goals underpin policy interventions specifically addressing loneparent households (Table 1.7). Preventing and alleviating poverty for the household, and especially for children, is one of the most recurrent goals. This is done to buffer the negative consequences of living in a single-earner household. The most popular


**Table 1.7** Policies specifically supporting lone parents by aim and OECD country

Source: Elaborated from OECD (2011, p. 219)

a Except for France, Iceland, Ireland, and Japan

policy tools in this case are tax breaks, family allowances supplements, advances of maintenance payments, childcare benefits, social assistance or housing supplements, and income supplements for lone parents (OECD 2011). Allowing parents to stay in the labour market (see Brady in this volume) is also an exigency that policies attempt to respond to in order to buffer the consequences of the double-burden dilemma.3 Some of these policies primarily intend to support lone parents with their caring commitments by either subsidizing alternative care (i.e. childcare benefits) or providing paid leave so they can take time to undertake the care themselves (i.e. parental leaves). Other policies seek to help parents stay in the labour market through workfare programmes attached to income support payments. Such policies are generally referred to as active labour market policies (ALMP) (Martin and

<sup>3</sup>Lone parents are simultaneously single earners and single carers in the household.

Millar 2003), as it is clear the same programme can serve more than one policy aim (e.g., income support to alleviate poverty and improve labour market attachment).

Researchers started to investigate the impact of social policies on lone parents' outcomes, arriving at two important conclusions: First, assessing the overall policy framework and the interaction among policies is important to have a better understanding of the outcomes for lone parents (e.g., Misra et al. 2007); second, the role of universalistic interventions in buffering lone parenthood negative outcomes is not to be underestimated, as shown by some studies (Brady and Burroway 2010; OECD 2011).

#### **Overview of the Book**

The purpose of this book is to approach lone parenthood from a life course perspective. The contributed chapters address critical aspects of the life course perspective like the interdependence of multiple life domains, linked lives, and the relevance of individual agency in context. They are organized in order to cover major issues at stake in the study of lone parenthood and lone parents' life courses. The first part of the book, including four chapters, is devoted to the definition of the phenomenon on lone parenthood, whether it is related to the ways in which lone parents are identified in the statistical and administrative records, the public debate, or the comparison between such external definitions and the more subjective self-image of lone parents themselves.

In the opening Chap. 2, Letablier and Wall broadly discuss lone parenthood in the context of family diversity characterizing Europe since the 1960s. With a focus on France, they compare the prevalence and characteristics of lone parents through time and critically examine the criteria used to define lone parent families, based on residence or on economic responsibility for children. Central to this analysis is the discussion of the extent to which such criteria relate to concepts of parenthood and parental responsibilities and challenge current social policies schemes addressing poverty and deprivation risks incurred by children living in one-parent households.

Salter, in Chap. 3, touches on the important aspect of the public image on lone parents in the UK, a country with the highest percentage of lone mothers in Europe. She critically addresses the stereotypical social representation of lone mothers as economically dependent on public support and as poor-quality parents, through an analysis of newspaper articles that appeared during two specific years, 20 years apart, and belonging to opposite political perspectives. The results point out at a homogeneous discourse both across time and through the political spectrum.

When targeting lone parenthood, often studies are focussed on adult mothers and fathers. In Chap. 4, Portier-Le Cocq reports on a qualitative study among young teenage mothers in the UK. The chapter looks at their financial, emotional, and social ways of coping with their early motherhood. The study shows how young girls often do not get pregnant by choice but by force and have little or no access to emergency contraception. A wide range of reasons also prevent them from having an abortion, and many of them show regrets afterwards for having the baby. On the other hand, the teenage mothers showed a great urge to prove they could manage, including efforts to combine work and care. Poverty and dropping-out of school, however, often hamper these intentions.

Addressing the issue of definitions and concepts related to a changing phenomenon, Bernardi and Larenza, in Chap. 5, use a set of open-ended interviews with lone parents in Switzerland to capture the transition to lone parenthood as defined by those parents who experienced it. They question to what extent women and men living alone and having the sole responsibility for their children identify themselves as lone parents and as a consequence of what change in their life. They conclude that the transition to lone parenthood is an often non-linear and ambivalent one and give suggestions on how to improve current measurements of the timing and occurrence of lone parenthood.

The second part of the volume includes three chapters devoted to the analysis of the diversity of lone parents in term of their socio-demographic characteristics and their family trajectories and relationships.

Informal social support is the topic of Chap. 6, in which Keim employs a mixedmethods design to describe lone mothers' social network structure and composition as well as the ways in which such contacts contribute to mothers' well-being. Data consist of an original set of problem-centred interviews and structured network data of 26 unemployed lone mothers living in eastern Germany, a particularly vulnerable group of lone parents. The resulting network typologies account for the size, the composition, and the role of network members in providing support. Results highlight a large heterogeneity in the social integration of lone mothers and provide a solid description of the ways in which networks represent a resource for them.

Do lone mothers suffer from a double disadvantage on the labour market when they have a migrant background? Studies looking at the interaction between family structure and migration are still rare. In Chap. 7, Milewski, Struffolino, and Bernardi analyse the Swiss Labour Force survey to examine the employment situation of immigrant women, whether lone mothers or mothers in couple, and contrast them with natives. Thanks to the large dataset, they can define various employment statuses by differentiating between full-time, long and short part-time employment, and non-employment as well as migrant generation and origin. Results show the double risk of international migrant lone mothers to be either unemployed or work full-time more often than native lone mothers. In the Swiss context, where child care services are insufficient and the number of mothers who work full-time is relatively small, family work reconciliation may represent an additional disadvantage and a source of stress for migrant lone mothers, who often lack informal support from faraway family members.

In the third part of the volume, three chapters address two major research themes in studies of lone parenthood from a life course and social inequality perspective: Lone parents' relative and absolute poverty in different institutional contexts and how the transition to lone parenthood affects income trajectories.

Is a husband really the main protection from poverty for a (lone) mother? This question is answered in Chap. 8. Hübgen answers the question by taking a European perspective using pooled SILC data. Three hypotheses are tested: (1) gender inequalities in the labour market and the welfare system increase the poverty risk; (2) gender inequalities in working hours and wages increase the poverty risk, despite being employed; and (3) the less progressive work-family policies are in a country, the greater the young children increase lone mothers' poverty. The results show that gender inequality in a country is indeed responsible for part of the cross-country variation in lone mothers' poverty risk. It also turns out that the structure of the labour market is partly responsible for the extent to which full-time employment may have a poverty-reducing effect among lone mothers.

Mortelmans and Defever in Chap. 9 analyse the economic trajectories of those who enter lone parenthood after divorce, taking advantage of the Belgian register data for Flanders. They compare employment and family trajectories after marital breakups, with special attention to income losses and gains thereafter. Focusing on the differences between men and women, and between lone parents and other family structures, they conclude for women, and not for men, re-partnering represents a way to improve their financial status. Yet, all in all, the 'financial consequences of divorce are still highly gendered', with women who head lone-parent households suffering from longer-lasting income drops significantly more than men.

The British Household Panel Survey is an excellent data source to observe trends in families over a long period. In Chap. 10, Harkness uses the BHPS to get a closer understanding of the economic consequences of women making the transition to lone mothers. As other chapters in this book report for other European countries, UK lone mothers are faring badly economically and in terms of labour market participation. The chapter shows how these circumstances are to a large extent influenced by the occurrence of the lone motherhood in the life course. It is predominantly young and less educated women who suffer financially from lone motherhood. In addition, the low labour market participation of these women is not related to benefit receipt. Low labour market participation is associated with a selection into lone motherhood rather than a deliberate choice in order to receive state support.

The concluding four chapters (part IV of the volume) discuss the important interdependence on the one hand between lone parenthood and employment and, on the other hand, lone parenthood and health.

One of the main ways that many countries have sought to alleviate the poverty that many lone mothers face is through increasing their labour market attachment. In theory, earning a reasonable wage through paid employment should reduce lone mothers' poverty, but in practice, this is not always the case due to high childcare costs and the withdrawal of income support payments once these individuals start earning an income. In Chap. 11, Brady draws on data from a three-year qualitative longitudinal study to illustrate how access to informal childcare in the context of a mixed (formal and informal) childcare package can be a crucial factor in enabling lone mothers to combine paid work and care in Australia. The study shows that the combination of formal and informal childcare results in mothers being able to enter the labour force with greater ease, lessens the pressure they face in balancing paid work and care commitments, and helps them to adapt to their children's changing childcare needs. This assistance supports these lone mothers to have stronger labour force trajectories over time. This chapter illustrates that the childcare needs of lone mothers are very dynamic and much more complex than often assumed in quantitative studies. Furthermore, it shows that the inflexibility of formal childcare arrangements is far more problematic for lone mothers than previously assumed.

Struffolino and Mortelmans, in Chap. 12, start with the fact that the poverty risk for one-parent families is significantly higher compared to households composed of couples with children and compared to the general population. The authors analyse data from the Crossroads Bank of Social Security (CBSS Datawarehouse) and study the Flemish lone mothers' patterns of labour market participation. Given a context of generous unemployment and that social assistance benefits are available, their aim is to establish what individual and household characteristics of lone parents are associated with which kind of employment trajectories. It appears that an early selection into lone motherhood is associated with lower employment opportunities. Additionally, the number of children underage living in the household negatively influences the labour market attachment of lone mothers and raises their likelihood of being unemployed or on social benefits.

Selection mechanisms are at the core of the investigation of the association of subjective health and lone parenthood in Chap. 13 by Dinescu, Haney-Claus2 , Turkheimer, and Emery. Adopting a twin research approach, and the large Swedish Twin Register dataset, they determine the extent to which the observed relation between lone parenthood and depression is the result of genetic and shared environmental confounds. The authors show that despite selection being responsible for part of the association between lone motherhood and depression, it does not rule out the presence of causal mechanisms related to lone parenthood itself.

A longitudinal setup characterizes a complementary study by Kuhn in Chap. 14, which looks at changes in satisfaction with health and well-being of lone mothers during the transition into lone motherhood. The results unequivocally show that, besides separation, the transition into and the duration of lone motherhood have significant negative effects on health satisfaction and well-being. Yet, there are important buffering factors such as being in a LAT relationship and the number of underage children. Socioeconomic circumstances, such as income change and employment status, play an important moderation and result in heterogeneity in well-being trajectories during the transition into lone motherhood.

Selection is important to understand social phenomena. The possible neglect of selection bias is also one of the drawbacks of many studies on the consequences of divorces. In Chap. 15, Hannan presents empirical evidence for the association between lone parenthood and children's well-being. Using the Growing Up in Ireland child cohort study, the results show that despite the huge selection differences into family types, the structure of the family does have an influence on children's well-being. More specifically, math scores and school attendance remain lower among children from never-married one-parent families. Nevertheless, taking out the selection effects drastically reduced the differences between families, indicating that it is not the lone parenthood status as such that predominantly determines child outcomes.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **References**


**Laura Bernardi** is Professor of Life Course Demography and Sociology at the Faculty of Social and Political Sciences at the University of Lausanne, Switzerland, and Deputy Director of the Swiss National Center for Competence in Research LIVES. Her major research areas are fertility, family demography, migration and life course. She has published in major demograpic and sociological journals on topics like fertility intentions and behaviour, social network influences on family development, intergenerational transmission of fertility and family formation, lone parenthood and health, migration and family formation. She is also co-editing the Springer series on Life Course and Social Policies.

**Dimitri Mortelmans** is Full Professor of Sociology at the Faculty of Political and Social Sciences of the University of Antwerp, Belgium. He teaches qualitative research methods, applied multivariate statistics and advanced topics in family sociology, life-course sociology and demography. He is head of the Centre for Longitudinal and Life Course Studies (CELLO). His research concentrates on family sociology and sociology of labour. He has published on divorce, new constituted families, gendered labour careers and work–life balance. He is also the main author of the Step in Statistics book series of which six volumes have been published (in Dutch). On qualitative methodology, he published the Handbook of Qualitative Research Methods and Qualitative Analysis with Nvivo.

**Ornella Larenza** is a Social Sciences PhD student at the University of Lausanne. She works at the National Centre for Competence in Research LIVES at the University of Lausanne where she is involved in a qualitative panel research on lone parenthood in French speaking Switzerland. Her main areas of research are family sociology and policy, care, social policy and life course.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Part I Defining Lone Parents**

## **Chapter 2 Changing Lone Parenthood Patterns: New Challenges for Policy and Research**

**Marie-Thérèse Letablier and Karin Wall**

Lone parenthood has become a widespread family living arrangement in the Western world. Not only has the number of lone-parent families been growing, but their profiles have changed substantially. Until the mid-1990s, lone parenthood was mainly the outcome of two trajectories: dissolution of a union due to the death of one parent, and childbirth out of wedlock. These two trajectories shaped at least two lone-parenthood patterns, which were viewed differently according to traditional social representations of family, filiation and parenthood. While widowed lone parents received compassion from society, unmarried mothers were stigmatized on account of their 'immoral behaviour'. Accordingly, their offspring were also perceived differently: compassion and solicitousness was accorded to orphans whereas shame awaited the children of unmarried mothers.

Today, in most European countries, lone parenthood results primarily from the dissolution of the matrimonial bond or the breakdown of a cohabiting relationship. This means that many children in lone-parent families have two parents but do not live full time with both of them under the same roof. Not only has change in lone parenting patterns modified the perception of these families but it has also contributed to the emergence of new questions for society. Policy objectives no longer focus solely on the replacement of the economic provider, generally the father, but extend to the maintenance of parental ties after the separation of the couple and on the risk of social exclusion for poor lone parents.

New patterns of lone parenthood present a number of challenges both for social policy and family research. First, lone parenthood has become more difficult to identify due to the greater diversity and complexity of living arrangements.

M.-T. Letablier (\*)

K. Wall

Centre national de la recherche scientifique – CNRS – and Centre d'Economie de la Sorbonne – CES – Université Paris1 Panthéon-Sorbonne, Paris, France e-mail: marie-therese.letablier@univ-paris1.fr

Institute for Social Sciences, University of Lisbon, Lisbon, Portugal

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_2

Secondly, lone parenthood is often a transitional step, since most lone parents are likely to repartner and reconstitute a family. The duration of lone parenthood varies according to sex, age and number of children. It should therefore be seen as a transitional period in the life cycle. While separated parents are likely to find a new partner and found a new family, children find a new place among an increasingly heterogeneous group of siblings, composed of children from different unions. Reconstituted families call into question traditional concepts of kinship and parenthood that are no longer restricted to biological ties. For a growing number of children, the term 'multi-parenthood' provides a more accurate representation of the diversity and complexity of family ties. Thirdly, lone-parent families deserve special attention in social policy because of their higher poverty risk.

In this chapter, we explore how lone parenthood challenges both family sociology and social policy in European countries, and more specifically in France, by focusing on three issues: first, the problem of defining lone parenthood, especially in view of the increasing complexity of family living arrangements; secondly, the prevalence of lone parent families and the relevance of a life-course perspective for capturing the changing patterns of lone parenthood; and, thirdly, the role of social policy in shaping lone parenthood.

#### **A Fuzzy Definition of Lone Parenthood**

Although there is a large body of research on lone parenthood, there have been relatively few cross-country comparative analyses of lone parent families. One reason is probably the lack of a common definition of lone parenthood for all countries in Europe, in addition to differences in national data sources. Another reason is the difficulty in reaching a precise definition of the concept. The definition that is usually used in the comparative literature tends to consider lone parent families as those "where a parent lives with his/her dependent children, without a spouse/partner, either on their own or in multi-unit households" (Duncan and Edwards 1997). This apparent simplicity in the definition conceals a number of problems that are amplified in cross-national comparisons. Three main issues may be mentioned: the age limit of children who are 'living with' the parent (under age 15, 16, 18 or 24); the criteria used to define a 'lone' living arrangement (which may include parents living on their own as well as those living with other adults in multi-unit households); and the meaning of being 'without a partner', in particular in a post-divorce/separation family arrangement. In the latter case, children may spend all their time with one parent but they may also spend time with the other parent or even alternate between two separate homes; moreover, the 'non-residential' parent may be involved in both the economic maintenance and the upbringing of his/her children, thereby contradicting the idea of a lone parent who is the only breadwinner and caregiver of his/her children.

Dependent children are not defined similarly in all countries and in all surveys. The age limit applied to children varies according to the data source and across countries depending on the norms for school-leaving age and the age of emancipation from the family. In the European Labour Force Survey (ELFS), a lone-parent family includes at least one child below the age of 15, whereas in the European Community Household Panel (ECHP) a lone-parent family includes at least one child below the age of 16 or a child aged from 16 to 24 who is 'economically inactive', implying that s/he is economically dependent on his/her parent/s. In the European Panel Survey on Income and Living Conditions (EU-SILC), a lone-parent family includes at least one child below the age of 17 or an economically inactive child aged between 17 and 24. Similarly, in France, the age limit for children varies according to the data source. While data produced by the French *Institut national de la statistique et des études économiques (INSEE)* consider children or young adults under 25 who are economically dependent on their parents, the *Caisse nationale d'allocations familiales (CNAF)* usually refers to children under 20. Whereas INSEE gives preference to the 'residence' criterion (living with), CNAF emphasizes the 'budget unit' (household income and expenditure). The age criterion roughly reflects the degree of dependency, and therefore produces different assessments of the number of lone-parent families, even for the same year in the same country. The concept of dependent children is a social construction that depends on the purpose of the data collection as well as the institutional and cultural context. It partly reflects how children /young people are treated in the benefit system. Moreover, "dependence" covers financial dependence as well as dependence for care and other needs. So age matters, since childcare for instance is of little relevance for adult children.

Lone parents living in multi-unit households are particularly difficult to detect, especially when the household contains three generations. Moreover, they may or may not find parental substitutes and strong support in grandparents or other adults living in the same household, making it difficult to assess whether these are 'lone parent', 'multi-parent' or 'extended lone parent' families. The 'living alone' criterion is therefore controversial. Does it mean "living on one's own" and bringing up children without any economic or parenting support from other "breadwinners" or does it mean "living out of a stable partnership"? In most surveys it is understood as "not living with the other parent, nor with a new partner" though the "lone parent" may be living in a multi-unit household. In other words, the concept seems to draw specifically on the model of the 'nuclear family household' (cohabiting couple with children): lone parents are those who do not cohabit with a partner, whether or not they have support from other significant persons, be they non-residential partners, ex-partners or others.

Social and family policy institutions also adopt the 'living alone without a partner' criterion to exclude from support lone parents who are involved with a new partner even if the relationship is unstable. The criterion is based on a subjective statement by an individual, which is particularly problematic for lone parents in French overseas territories where lone parenthood is a widespread social phenomenon1 and where the relationships between all household members are not clearly

<sup>1</sup> In Martinique, 53% of households with dependent children are lone-parent households, and six out of ten teenagers live partly or wholly with one parent for the first 15 years of life.

captured in surveys. In the French Caribbean territories where one-parent families are twice as numerous as in continental France, the lone parent is almost always the mother but in most cases fathers do not completely disappear. In 2004, approximately 70% of childbirths were out of wedlock in Martinique, often involving 'multi-partnership' for fathers (Lefaucheur and Brown 2011). Anthropologists have shown that the extension of these family forms reveals the centrality of the maternal figure, while the father's role is blurred (Mulot 2013). This example shows that, in some societies, lone motherhood is embedded in cultural and social gender conventions. The traditional nuclear family composed of a parental couple and children throughout life, hitherto considered as the reference norm in European countries, is no longer relevant for the variety of family life arrangements in a global context of growing cultural diversity.

Lastly, the concept of lone parenthood is not always appropriate because a growing number of children who live in lone-parent families have two parents. In the 1970s, at least in France, the issue in public debate was how to legitimate lone parents as families while avoiding the stigmatization of unmarried mothers (Lefaucheur 1986). In the twenty-first century, the issue is how to identify families in which children do not live with their two parents, in the knowledge that many children living mainly with one parent have a more or less regular relationship with the nonresidential parent who contributes to their maintenance and upbringing. The concept is made more complex by the need to distinguish lone-parent families in which children never see the non-residential parent (a "lone" lone-parent family) from those where children regularly stay with the non-residential parent and/or share their time and residence between the two parents (a lone "bi-parental" family). In this case, the child can be said to live in two households with two parents who share maintenance, care, education and love. From a sociological perspective, it is essential to identify and conceptualize these situations as different. Moreover, the development of shared residence and shared parenting in some countries raises the problem of the double counting of lone-parent families. This is a crucial issue for household surveys to deal with (Toulemon and Pennec 2010).

Research on families in other societies also points to the need to reflect more carefully on the appropriateness of the concept of lone parenthood. Authors of a survey of conjugal violence against women in Martinique distinguish three patterns of lone motherhood: lone mothers (23%) who have never been married or cohabited with a partner (more than 25% had always lived with their parents, even when aged over 40); lone mothers (35%) who have a regular relationship with the child's father who is partnered with another woman; and separated mothers (42%) whose children were born before the separation (Lefaucheur and Brown 2011). The term 'friends' was suggested by demographers in the beginning of the 1970s to identify the second type of lone mothers who have more or less regular sexual relationships with a partner with whom they are neither married nor cohabiting (Leridon et al. 1970). Anthropologists refer to this situation as a 'visiting union' in which fathers are 'visiting fathers' (Mulot 2013).

These examples suggest that lone parenthood takes place in a context of 'inconstance de la famille' ('changeable and plural family forms')2 (Leridon and Villeneuve-Gokalp 1994), and that lone-parent living arrangements are diverse and complex, hitherto difficult to capture. They also reveal that current definitions of lone parenthood are embedded in cultural norms emphasizing the concept of the nuclear family household based on the ideal of a cohabiting parental dyad. There has been little effort to unpack the concept itself and to grasp the diverse sociological meanings of a 'non-cohabiting mother/father living with dependent children'. These problems and gaps would call for more research as well as new theoretical frame. For example, to move beyond the nuclear household concept, it would be useful to adopt network and configurational approaches to lone parenthood, in order to take into account family bonds and meanings that bridge across different generations, couples and households, as several authors suggest (Widmer 2010; Wall and Gouveia 2014). A life course perspective, bringing in the principle of linked or interdependent lives, would also provide a more dynamic and complex view of lone parenthood.

#### **Variation and Change in the Prevalence of Lone Parenthood**

The increase in divorce and separation involving children has led to an increase in lone parenthood over the last decades. The OECD Family database provides data on children's living arrangements, providing a comparative overview of the incidence of lone parenthood in Europe (Table 2.1).

The proportion of children below age 18 living with one parent varies extensively from 5.3% in Greece, 7.2% in Spain, 10.2% in Italy for the lowest rates to 24.3% in Ireland, 21.5% in the UK, 21.8% in Estonia for the highest rates in Europe. With 13.5% of children living with one parent, France stands in between, close to Austria, Finland, and Germany (Table 2.1). The lowest shares are found in Southern European countries whereas the highest are found in Nordic countries and above all in the UK and Ireland. With the exception of Estonia, the central and eastern European countries show close to average values, higher in Hungary and the Czech Republic than in Poland, Slovak Republic and Slovenia.

Although the OECD does not provide data on different types of "one parent families", it informs on the share of children who live with two married parents (Table 2.1). Again, the proportion is far higher in Southern European countries than in the Nordic countries or in the UK. The highest shares are found in Greece, Italy, Spain, while the lowest are found in Sweden and Estonia. Moreover, the proportion of children living in multigenerational households varies from less than 1% in Sweden, the Netherlands, Denmark, Finland and Germany, to more than 10% in Hungary, Portugal, and Estonia. The highest rate is found in Poland, where 22% of

<sup>2</sup>This play on words refers to the "inconstance" (inconstancy) in love relationships and applies it to change in family living arrangements. It also refers to instability of the couple over the life course.


**Table 2.1** Distribution of children by household type in selected EU countries, 2007

Notes: 1- The category "cohabiting parents" includes unmarried parents and parents in reconstituted households. 2- Children are defined as household members under 18 Source: OECD, Doing better for families, 2011 (OECD family data base)

children below the age of 18 live in such households. Data do not however specify whether these households include the two parents of the child or one only, or even no parents. These figures reflect the contrasted incidence of intergenerational coresidence and support, which tends to be stronger in Central, Eastern and Southern European countries than in Scandinavia, where individual and residential autonomy is less of a problem.

The prevalence of lone parent families must therefore be examined in the wider context of the increasing diversity and complexity of family living arrangements, norms and practices. It should also be examined in a life course perspective given the changing family forms over the life course due to breaks and discontinuities in couples' life. The lowest rate of lone parenthood seems to be correlated with the high incidence of marriage and subsequently with the lower incidence of unmarried partnership as observed in Greece, Spain and Italy (Table 2.1). It reflects a family pattern in which marriage remains the norm for couple's living arrangements, unmarried cohabitation is low, divorce is not yet widespread and intergenerational family solidarity is more common. This pattern contrasts with a family pattern characterized by more fluidity and diversity of family forms, a relatively high proportion of unmarried parents, a high rate of divorce and a high proportion of children living either with one parent for most of the time, or in a reconstituted family, or alternately with each of the parents. This pattern in which very few children live in multi-generational households is commonly widespread in Nordic countries. In between are different sub-patterns combining a low or medium rate of children living with a lone parent with various shares of children living with two cohabiting parents or in reconstituted families.

In summary, a significant number of children now live most or all of the time with one of their parents, and the increase in the proportion of divorces involving children has been accompanied by a rise in lone parenthood over the past few decades (Chapple 2009; Chzhen and Bradshaw 2012) Across the OECD countries, one child in eight (15%) lives with one parent only, while nearly 84% of children live with two married or cohabiting parents. Moreover, projections to 2025–2030 suggest that numbers are likely to increase in many countries, even if the effect of a rise in divorce and separation is unlikely to substantially mitigate that of declining numbers of children (OECD 2011). Beyond this general trend, the changing diversity and complexity of lone parent families has to be examined in the context of multiple shaping factors such as changing gender and parental roles, marital status, trends to repartner, number and age of children, and lone parents' employment and living conditions. As mentioned above, it is also crucial to introduce a life course perspective in order to understand the duration and incidence of lone parenthood across different life stages.

Much comparative cross-national data is missing on these issues, and it is mainly national data sets that allow us to highlight and fully understand the main trends related to the new plurality and complexity of contemporary forms of lone parenthood, in particular in a life course perspective. We will therefore begin by analyzing the French case before looking again at lone parenthood patterns in comparative perspective.

#### *Changing Reasons and Trajectories of Lone Parenthood in France*

Until the 1970s, lone parenthood in France resulted predominantly from the death of one of the parents and/or from single3 motherhood. Today, lone parenthood results mainly from divorce or separation of a cohabiting couple. More than 62% of lone parents were previously married or cohabiting while 20% never lived in a couple and 18% became lone parents after the death of a spouse or partner (Letablier 2011a).

Lone parenthood due to the death of one parent is therefore much less likely to be the reason for 'living alone with children' in France. The proportion of widows among lone parents fell from 55% in the early 1960s to under 10% in the mid-2000s (Chardon et al. 2008). However, the high incidence of accidents at work among

<sup>3</sup>The term 'single' is used here in the sense of never married or partnered, by contrast to 'lone' that refers to living without the father or a new partner.

manual workers explains the larger number of orphans among children of blue collar workers: 9.8% compared to 5.1% for children of professional workers. On average, widowed lone parents are older and better off than other lone parents. The majority are mothers, with children on average older than in other lone-parent families. Lone parenthood due to the death of one of the parents is however more often temporary than for other forms of lone parenting. A French survey of family histories carried out since the 1990s by the *Institut national d'études démographiques* (INED) shows that 50% of widowed lone parents were partnered again 4 or 5 years after the death of the partner, whereas 25% were still living alone 19 years later, mostly women with several children. The survey also shows that widowers were more likely to repartner than widows, all the more so if they are women with several children (Lefèvre and Fillon 2005).

Single mothers who never married or cohabited before childbirth display a second profile of lone parenthood, which varies according to different life trajectories. Some result from the choice of having a child with a married man for instance, or the mother's decision to adopt or get pregnant and bring up a child by herself, while others result from an unwanted pregnancy which was further accepted. Although it is difficult to assess the number of single mothers who have not lived with a partner before childbirth or the number of separations during pregnancy, INED estimates at between 5% and 7% per year the proportion of such births in France. The proportion of unpartnered or unmarried mothers accounts for around 15% of all lone parents, including 13% of mothers without any form of cohabitation and 2% of couples that have separated before birth (Haut Conseil de la Famille – HCF 2010). These never partnered mothers are generally younger than lone mothers in general. They also have a lower level of education, but in contrast to the UK and Ireland where early pregnancies are more widespread, they are rather exceptional in France.

A third profile refers to couples who have separated or divorced. This life-course trajectory now represents the most widespread reason for lone parenthood in most European countries. It is however uneasy to distinguish lone parenthood resulting from the separation of a never-married couple from that due to the divorce of a married couple. Estimates for France suggest that half of relationship breakdowns are due to divorce and half to separations of never-married couples. Divorce occurs on average after 14 years of marriage, at the age of 40 for women and 43 for men. Altogether 56% of divorces involve at least one dependent child. Re-partnering after divorce or separation occurs more or less rapidly according to parental status, custody arrangements and number and age of children. On average, half of divorced parents re-partner within 5 years after divorce, compared to half of separated parents who re-partner within 3 years after separation. Divorced parents are on average older than separated parents, and their children are also older: eight and a half compared to four and a half. With regard to the level of education and qualifications, divorced or separated mothers do not differ from partnered mothers. They have a higher level of education and higher labour force participation than never partnered lone mothers.

Post-divorce and post-separation custody arrangements are also important to understand the transitions into and out of lone parenthood. According to tax declarations, 1 year after the divorce of the parents, 76% of dependent children in France

live most of the time with the mother, only 9% with the father and 15% alternately with mothers and fathers (Bonnet et al. 2015). This is different from Sweden, where only 65% of dependent children with separated parents live most of the time with their mother, 17% with their father, and 18% share their time equally between the two parents (Hobson et al. 2011). In contrast, in the UK the vast majority of children live with their mother (Finn 2011). These figures reflect the social acceptability of lone parenthood as well as the changing perceptions of gender roles and parental responsibilities. For example, whereas the focus in Sweden is on children and their living arrangements according to a child-centred perspective, more attention has been paid to lone mothers in the UK (mother-centred approach), while the focus was for a long time on families (family-centred) in France (Letablier 2011b). Whatever the approach, living arrangements for children are increasingly complex and uneasy to capture, depending mainly on the amount of time they spend respectively in the mother's or father's home, as well as on each parent's economic contribution to the child maintenance, education and care. In addition, as mentioned above, these living arrangements may change over time since the lone parent may find a new partner and children may want to change their place of residence as they grow up.

The social and ethnic status is another variable that shapes transitions to parenthood and lone parenthood. The number of lone parents has grown more rapidly among immigrants than among the native French population. Their number more than doubled over the last two decades, from 143,000 in 1990 to 298,000 in 2008 (Mainguené 2013). On average, 6% of immigrants aged 18 and over are lone parents compared to 4% of non-immigrants. The increase is concomitant with the changing migration regime. The "new" regime differs from the previous one regarding origin, gender, and objectives. Each of these features may affect lone parenthood growth among migrants. First, with respect to regional origin, although the share of lone parents is roughly similar to the average for immigrants from Europe, Morocco, Tunisia or Turkey, it is notably higher for immigrants from other African regions. The proportion of lone parents among the "new migrants" from sub-Saharan Africa and Asia is as high as 12% (Insee 2012). One explanation could be the social and family structures in the country of origin. In the Sahel, men, women and children generally live separately in different places of residence. This contrasts with the Western conception of the family as a unit composed of a married/partnered couple living with their children in the same home. The second change in the migration regime results from the growing number of women among the new migrants. Thirdly, women are now not only more numerous than men among migrants but they also do not always come to join a husband or a partner but instead have their own autonomous life project. Nowadays, the main reason for coming to France is to study or to work (Beauchemin et al. 2013). Most of these new female migrants come from the Ivory Coast, Cameroun or the Democratic Republic of Congo. Some of them are not married while others have left their husbands in the country of origin. These 'living apart together families' or "distance families" subjectively feel they are lone parents. The more rapid growth of lone parents among immigrant families than among the French native population suggests that changes in the migratory regime and the experience of lone parenthood might be linked. So far, the issue has not been fully explored.

Changes in the reasons for lone parenthood, together with the decline in female adult mortality and the fact that, after divorce or separation, the vast majority of children live mainly with their mother, explain the low proportion of lone fathers among lone parents. The share of lone fathers in France fell from 20% in 1970 to 14% in 1990, then rose to 15% in 2005 (Chardon et al. 2008). The rebound in the mid-2000s reflects the emergence of a new lone parenthood pattern in which fathers are more involved in parental responsibilities and duties than they were previously. As mentioned above, this also results from a (slow) change in custodial arrangements since alternating custody now concerns 15% of children 1 year after the divorce of their parents (Bonnet et al. 2015). It also reflects a change in men's aspirations to share parental responsibilities, as well as policy incentives to achieve gender equality in family life. The French law of 1970 on joint parental responsibility legally and symbolically closed the era of the *Pater familias* as the economic provider for the family and main source of authority. Then the law on joint parental responsibility in 2002 set forth the principle of co-parenting according to which parental responsibilities should be equally shared between the parents, whatever their living arrangements. Concomitantly, societal expectations regarding fatherhood gradually changed. Fathers of today are expected to share the responsibility for childrearing with mothers, and to be increasingly involved in their children's lives, whatever their living arrangements with the mother. While mothers' participation in the labour force has been continuously increasing over the past 50 years, reconciliation of work and family life has become an issue for both mothers and fathers. After divorce, an increasing number of fathers apply for shared custody, making lone fathers an issue of growing interest for research (Martial 2013a, b). The relationship with the non-residential parent is found to depend on the father's involvement in childcare before separation. Fathers who had been sharing parental responsibilities before splitting up often claim equal sharing after separation. The success of co-parenting also depends on the quality of divorce measured in terms of the level of conflict before and during the procedure (Martial 2013c).

#### **Lone Parenthood in France as a Transition in the Family Life Course**

Surveys usually capture living arrangements at a particular point in time, thereby providing a snapshot of lone parenthood, rather than tracking changes over the life course. Given the change in lone parenthood patterns, with divorce and separation frequently followed by re-partnering and family reconstitution, it is ever more relevant to conceptualize and analyze lone parenthood as a stage in the life course. Today, in France, 11% of children live in a reconstituted family with brothers and sisters from a previous union. Reconstituted families account for almost 8% of families with at least one child under 18. One marriage in five is a remarriage, and 93% of remarriages occur after a divorce while 7% occur after the death of one of the parents.

Repartnering after the dissolution of a marriage or after separation of a cohabiting couple depends on factors which not only weigh on the likelihood of the repartnering process but also on its tempo. The presence of children from a prior union and the custody arrangements affects the repartnering process (Vanassche et al. 2015). Interestingly, a study based on survey data in Flanders shows that compared with other divorcees, full time residential parents are the least likely to start a new union following a separation, and that parents are more likely to start a union with another parent than with a childless partner. Evidence from this study suggests that parenthood may not be a particularly attractive status for post–separation union formation. However, results from French surveys show that parental status has a different incidence for men and women, as well as the living arrangements prior to lone parenthood. After a divorce, fathers who have custody of their children reconstitute a family more rapidly than fathers without custody: more than one in two fathers with custody are re-partnered 4 years after the union breakdown, compared to 39% of fathers without custody. The reverse applies to women, but the gap is smaller than for fathers. Women re-partner more rapidly when they do not have custody: one third of mothers without custody will again be living in a couple 4 years after the breakdown, compared to 27% of those who have custody These figures suggest that the custody arrangements strongly influence the process of repartnering. They also suggest that the effects of the child "burden" on new relationships differ for men and women.

Whereas the duration of lone parenthood is on average 5.5 years, the duration is lower for separated parents (4.5 years on average) or widowed (5.5 years) and higher for never partnered parents (10 years). The duration is also longer for lone mothers than for lone fathers, respectively 5.6 years and 4.2 years on average, but the gap is partly due to the fact that one lone mother in six has never been partnered whereas this situation is rare for lone fathers. Duration of lone parenthood is also longer for low educated parents than for other lone parents: 24% of lone mothers and 15% of lone fathers with a low level of education have been lone parents for more than 10 years compared to 14% of lone mothers and 7% of lone fathers with a university degree (Buisson et al. 2015). Therefore, men are more likely than women to reconstitute a family and to do so more quickly. Reconstituting a family is three times more common for widowed men than for widowed women aged below 55: about half of widowed men below 55 re-partnered after the death of their wife, compared with only 17% of widowed women, from 1990 onwards (Lefèvre and Filhon 2005).

Repartnering and childbearing produce increasingly complex family patterns. The concept of 'pluri-parentality' is used to capture the various relationships between children and their biological or social parents (Théry 2013). For instance, the biological father may not be the parent who looks after the child in everyday life, and the child may have closer bonds with the stepfather than with the biological father, thereby raising new issues for family regulation. Current controversies in France about granting stepfathers a legal status testify to the tensions surrounding the conceptualization of parenthood and parenting.4

#### *Contrasting Patterns of Lone Parenthood in European Societies*

Changes in the profiles of lone parents reflect a move from a lone parenthood regime in which one of the parents is missing to a regime in which high proportions of children have two parents who live separately and who may or may not have been married or living together as partners. This change over time has not been uniform in all countries, depending on attitudes and policies towards marriage and divorce, as well as on the representations of what the family is or should be. So the three main life trajectories leading to lone parenthood are not distributed in the same way in all European countries. Comparative data provided by the EU Statistics on Income and Living conditions (Chzen and Bradshaw 2012) and a comparative study carried out for the European Commission by the Brodolini Foundation (Trifiletti 2007) show that the vast majority of lone parents are divorced, separated or single parents, whereas widowed lone parents have become a minority in all EU countries, in spite of some cross-national variations. In 2009, proportions of widowed lone parents were higher in former socialist countries (24% in Romania, 12–14% in Poland, Lithuania, Bulgaria and Hungary), as well as in Greece (15%), Cyprus (15%), Italy (11%) and Spain (11%); in other countries proportions of widowed parents were as low as 2–3% in Sweden, UK, Ireland, Denmark, Finland and Iceland, and below 11% in the other countries.

The other two trajectories into lone parenthood also show some striking crossnational variations. Three patterns may be highlighted. A first pattern linked to high and almost equal proportions of both never-partnered and divorced/separated lone parents. In the UK, Ireland, Sweden, Norway, Iceland, and Slovenia at least half of all lone parents report their marital status as single, thereby highlighting not only the well-known impact of early pregnancies, in particular in the UK and Ireland, but also a possibly growing trajectory linked to voluntary independent motherhood or deriving from "living apart" couples. Sweden stands out among Scandinavian countries due to its relatively high age of lone parenthood: only 18% of lone parents are under age 35, compared to 38% in the UK. Routes to lone parenthood differ in these two countries, especially with respect to the incidence of early pregnancies, which are not as common in Sweden. Lone parenthood is widespread in Ireland and the UK, due in particular to young un-partnered mothers. The high proportion of early pregnancies in these two countries is often associated with lone motherhood, though

<sup>4</sup>The introduction of a bill on 'parental responsibility and child's interest' was postponed because of strong opposition. The draft discussed in Parliament in 2014 defined the legal status of stepparents, thereby acknowledging the value of the relationship between children and step-parents. Opponents advocated in favor of a sole type of parenthood based on biological links rather than on everyday care relationships.

not always. As a consequence, the average age of lone mothers is relatively low: 26% are under 30 in the UK compared to 12% of mothers living with a partner. More often too, they live in social housing or private rented accommodation, and their income is low (Finn 2011).

A second contrasting pattern is associated with a relatively low proportion (around one quarter or below) of single lone parents, mainly in new accession countries and some southern European countries (e.g. Romania, Hungary, Slovakia, Greece, Cyprus, Italy) and higher proportions of divorced (rather than separated) lone parents; as mentioned above, this pattern emerges in social contexts where marriage is still more widespread than cohabitation.

Finally, a third pattern reveals average proportions (around one third or slightly above) of single lone parents and varying proportions of both separated and divorced lone parents. In continental central European countries this would seem to represent the mainstream pattern, with trajectories linked to separation and divorce representing the predominant trend, and single un-partnered mothers a persistent, rather than a declining, minority trend. Given that in these countries early pregnancies are rarer than in the UK and Ireland, the un-partnered pathway into lone parenthood may indicate a new pluralization of single motherhood and fatherhood, including both voluntary and involuntary transitions to parenthood.

As mentioned above, these patterns have to be examined in the context of other crucial variables such as living arrangements, gender, age and re-partnering. There is a striking divide between the old and new accession states in the share of multiunit households as opposed to living on their own. In some countries lone parents do not live alone with their children but with grandparents or in multi-unit households: in many former socialist and some southern European countries this involves the majority of lone parents (over half of all lone parents in Romania, Bulgaria, Hungary, Poland, Latvia, Slovakia, Slovenia, Greece, and Spain), pointing to vulnerable living conditions, difficult or late access to residential autonomy, and longstanding practices of multigenerational co-residence. National data in some countries show that it is more likely to find single, young and low educated lone parents, rather than divorced lone parents (usually older and better employed/educated), in this situation. By contrast, the share of lone parents living in multi-unit households in most EU countries is much lower (less than one third) but still represents a relevant trend that must be taken into account.

Within these broad patterns of lone parenthood combining types of trajectories into lone parenthood and living arrangements, lone-mother families continue to predominate, with estimates from the Luxembourg Income study showing that only 11.7% of one parent families are headed by fathers, but with significant national differences. Proportions are as high as 19% in Sweden or Norway or as low as 8% in the UK or Spain or even 5% in Germany. A new profile of lone-father families may thus be gathering strength in some national contexts, in particular those with policies and practices more focused on gender equality in parenting, in response to involved fatherhood and more equal or shared gender roles in caring for young children. On average, lone fathers tend to be older than lone mothers.

Behind these contrasted patterns are not only changing norms and practices related to marriage and parenthood, but also increasingly complex forms of family formation and reconstitution. To develop and explore these patterns, a life course perspective allowing to capture lone parent trajectories and meanings over time and across countries would greatly enhance our understanding of the current plurality of lone parenthood. For the moment, we can only rely on national approaches to this issue. National datasets reveal that lone parenthood is a temporary situation for many lone parents. Lone parenthood due to the death of a parent is often more temporary than for other trajectories to lone parenthood, with most widowed persons having repartnered after 4 or 5 years; however, this transition out of widowed lone parenthood seems to be slower and more difficult in Central and Southern European countries, probably due to the fact that older children leave home late. Separated parents also tend to repartner more rapidly than divorced lone parents, while neverpartnered lone parents, especially those with low educational levels or with several children, are those that tend to repartner less or less rapidly. However, cross-national comparisons related to these pathways are still to be explored.

In sum, the existing comparative data reveal a plurality of lone parent trajectories and families across European societies. They allow us to (very broadly) identify some major cross-national patterns: a stronger focus on widowed and divorced (rather than separated) lone parents who also live more often in multiunit households in former socialist and southern European countries; a pattern strongly linked to early pregnancy within single lone parenthood as well as divorce and separation in the UK and Ireland; a pattern linked to high proportions of both single and separated/divorced lone parent trajectories and a strong increase in lone fathers in some of the gender-equality-oriented Nordic countries; a mainstream pattern in other central European countries such as France, Germany or the Netherlands, with somewhat lower proportions of single un-partnered lone parents and lone-father families, and a major focus on separated and divorced lone parents. However, the datasets also highlight huge gaps and insufficiencies in data analysis. We lack not only longitudinal datasets and in-depth national and cross-national studies of lone parent life trajectories but also simple cross-tabulations taking into account the characteristics of lone parents or multilevel analyses testing for the influence of different biographical and structural variables across different national contexts.

#### **Old and New Challenges for Family and Social Policy**

Lone parenthood challenges social policy, firstly by calling into question established parental norms, and secondly by exposing lone-parent families to the risk of poverty and social exclusion. Governments seek to develop policy measures in response to the issues raised by lone parenthood, and researchers are called upon to review their conceptual approach to lone parenting by taking account of changing patterns over the life course, the socio-demographic transformation of European societies, and their greater cultural diversity.

#### *Changing Norms: From Stigmatization for Moral Reasons to Stigmatization of Poverty and Welfare State Dependency*

Changing attitudes towards different family forms have accompanied changes in living arrangements. Unmarried cohabitation and divorce and the associated diversity in family trajectories are more readily accepted in society today, as testified by respondents in the French gender and generations survey (Mainguené 2011). Half of the respondents agreed that women can have and raise children alone if they wish, whereas respondents stress the importance of having two parents to bring up children.

The diversity of living arrangements is illustrated by the following statistics for France: in 2011, 72% of cohabiting couples were married, 22% were not married, 4% were in a civil partnership (Pacs),5 and 2% did not answer the question. Among the 0.5% recorded as living with a same-sex partner, a share of 16% was not daily living with the partner. Of these same-sex cohabitants, 60% were male and 40% female couples; while 43% were in a civil partnership, the rest were cohabiting (Buisson and Lapinte 2013). These figures reflect the complexity of living arrangements and the increasing disconnection between living under the same roof (co-residence) and being partnered or cohabiting (subjective perception of partnership). Figures also show that marriage is no longer a mandatory prelude to living as a couple and forming a family. Together with the diversification of family forms, the increase in separation has contributed to the trivialization of lone parenthood, which is no longer stigmatized for moral reasons: single mothers are no longer viewed as acting immorally and producing illegitimate children; and sexual relations outside wedlock are no longer viewed as reprehensible behaviour. Instead, stigmatization may be for welfare reasons: lone mothers may be stigmatized for their poverty and dependency on the welfare state, being seen as 'scroungers' always in need of support (Thane and Evans 2012). Discrimination may also emerge in other situations. For example, single women are not entitled to state support for assisted reproduction in all European countries, being accessible only for the heterosexual parental couple.

Attitudes have changed not only towards cohabitation, marriage and divorce, but also towards gender roles and parenthood. Fathers are likely to spend more time looking after their children, especially in the younger generations of fathers: 42% of the respondents in the Family and Housing survey carried out by INSEE said that 'it is neither better, nor worse, if a child stays with the mother after separation of the

<sup>5</sup>The *Pacte civil de solidarité-PACS* is one of two forms of union under the French Law. It is a partnership agreement between two same-sex or heterosexual adults. Since the creation of the Pacs in 1999 more than one million people have signed a partnership agreement. In 2011, 175,000 Pacs were contracted, compared to 250,000 marriages. Marriage was extended to same-sex couples in 2013, thereby opening the right for same-sex couples to adopt a child. By opening marriage to all couples, France was the ninth country in Europe to allow same-sex couples to marry, following the Netherlands, Belgium, Spain, Norway, Sweden, Portugal, Iceland and Denmark. In 2013, 7000 same-sex weddings were celebrated in France; in nine out of ten cases, the wedding involved two men (Bellamy and Beaumel 2014).

parents' while 30% assess that it is better for the child to live with the mother, and 28% disagree with this statement. Young respondents disagreed more often than older respondents with the systematic allocation of custody to the mother: 22% of those aged under 25 disagreed, compared to 45% of those over 75 (Buisson and Lapinte 2013).

#### *Policy Responses to Changing Models of Parenthood*

Divorce or separation of cohabiting parents challenge firstly the children's legal right to their two parents irrespective of their living arrangements, and secondly arrangements for ensuring that parental responsibilities including maintenance, education and childcare, are shared equitably between parents. Lone parenthood is seen as a risk for society, thus requiring a policy response: lone parenthood has become a major concern for governments and public policy because it is associated with the risk of poverty and social exclusion.

Despite the introduction of no-fault divorce in most European countries, conflicts often arise between parents about alimony, custody and the sharing of parental responsibilities. Conflicts arise also about the compensation for loss of income by a spouse who has devoted most of her life to looking after her family. In several countries, spousal maintenance has been progressively withdrawn as women's participation in the labour market has increased. Safeguarding the living standards of the spouse nevertheless remains a controversial issue for the courts. In France, the legislator has attempted to solve these conflicts by strengthening the role of family mediation, which is financially supported by the Family Allowances Fund. Divorce by mutual consent is now actively promoted and applies in more than half of all divorces in France. In 85% of cases, the court endorses the agreement reached by parents, while in 10% of cases it falls to the judge to make a decision about the child's residence, the level of alimony and custody.

The child's place of residence is another controversial issue. In all countries, most children live with their mother. In France, the child's main residence is assigned to the mother in 77% of divorce cases and in 84% of separations of nonmarried parents. Only 14% of children under 25 live most of the time with a lone father, usually older children: 18% of the 17–24 age group, compared to 10% of children under 6 and 3% of children under 2 (Chaussebourg and Baux 2007). Although custody is predominantly entrusted to the mother, shared-residence doubled between 2003 and 2012, from 12% to 21%. Shared residence is now determined in 30% of divorces by mutual consent. According to data published by the Department of Justice, altogether 17% of children with separated and divorced parents live alternately with fathers and mothers (Carrasco and Dufour 2015). The choice of residence depends however on the age of the child: children below six live more often with their mother, whereas grown up children, especially boys, tend to move to the father's residence. A consequence of growing shared residence is the decrease in the number of alimonies.

Encouraging fathers to share the responsibility for children more equally with mothers after separation has become a policy objective in several countries, although experts in child development and family associations do not always agree on the merits of such a decision, as shown by parliamentary debates in France over the new family law in 2014. On the one hand, proponents of the law, including associations of divorced and separated fathers, advocated alternating residence in the name of gender equality between parents; on the other, opponents argued for a traditional view of the family, whereby children should be cared for by their mother in the best interests of the child. Recent studies on fathers 'involvement in children's education and care identified three profiles of separate/divorced fathers, thereby illustrating changing contemporary fatherhood profiles. Changes have to do with policy discourse, public debates and practices as well. After having been 'absent' until the 1990s, fathers have become 'occasional' or 'intermittent' in the 2000s, then 'alternating' with mothers in shared/joint custody (Martial 2013a). The tempo of change differs across countries, depending on the policy and legal context, on the father's voice, and also on change in divorce regulations. The shared residence that has been legally encouraged in France for the last 15 years is conceived as a means of sharing parenting ("co-parentalité"). According to the French law, co-parenting requires a shared residence for children to spend the same amount of time with the two parents while having close relationships with the two parents. A comparison of shared residence arrangements in various European countries highlights the difficulties encountered in measuring these situations, in particular because there is no consensus on the terms used to describe these arrangements (Masardo 2014). The study also shows that the concept of shared residence in most European countries is not always effective in practice since the latter are shaped and limited by social norms regarding parenting and gender roles. Shared residence is however more common for children with high educated parents than for children whose parents have low educational attainment or are unemployed. The debate on shared residence is not closed; it includes issues relating to parenting, the respective roles of fathers and mothers, and what would be the best for the children. The lack of consensus within and between countries reflects contrasting views regarding the changing values underpinning the male breadwinner family model and, consequently, parental responsibilities and roles. Finally, although shared residence for children of separated parents has gradually increased, it is still more common in Nordic countries, especially in Sweden, than in South European countries where intergenerational solidarity is highly valued and where family members, particularly grandmothers, provide support for their sons as lone fathers. Qualitative studies of lone fatherhood suggest that this intergenerational arrangement is also common in France (Martial 2013b).

Decisions about child maintenance payments are made in 86% of divorce settlements in France, the exceptions being for parents who alternate custody. The level of payments is generally higher for fathers than for mothers, reflecting the gender income gap. A major concern for family policy is to support lone parents for whom maintenance is not paid or not paid regularly. Although the exact proportion of parents who do not pay alimony is not well reported, this situation is known to have a negative impact on living conditions for children. In some countries, temporary financial support is provided to lone parents while social services undertake proceedings to collect unpaid maintenance.

Another challenge for policy makers is how to ensure that parental contact is maintained after divorce. According to the French version of the *Gender and Generations Survey* carried out in 2005, approximately 15% of children living with a lone father had 'very little or no contact' with the mother during the year, compared to 40% of children living with a lone mother who rarely or never had contact with the father. However, 42% of children living mainly with their mother meet their father at least once a week (Chardon et al. 2008). So, children living with fathers have it easier to keep contacts with the mother than the reverse. On the one hand, there may be social norms on parenthood encouraging differential contacts by gender, while on the other hand, those fathers who have full custody are particularly selected as being 'good fathers' highly involved in the well-being of their children.

Another challenge for public policy is how to redefine the meaning of 'parenthood', especially when children are living in reconstituted or blended families. The increasing development of divorce and separation in family trajectories, together with repartnering and childbearing after separation has resulted in a complexification of family forms, as noted above,6 in which biological parents may not be the 'social' or main carer parent. The status of step-parents has become difficult to determine, since social practices change more rapidly than legislation (Lapinte 2013). In the 1990s, the French sociologist Irène Théry suggested using the term 'démariage' to denote a social phenomenon that was spreading rapidly at the end of the twentieth century (Théry 1993). The term refers less to the decline in marriage rates or to the depreciation of the value attributed to marriage than to the role of marriage as an institution in Western parenthood systems and, consequently, in the gendered organization of social life.

From being a widespread social norm until the 1970s, marriage has become a matter of personal choice. Rather than disqualifying marriage, the term 'démariage' relativizes it as being the starting point for establishing a home and a family. Today, both spousal and parental bonds are no longer regulated within the framework provided by the marriage institution. Since the 1972 French law establishing equality between children irrespective of filiation (married or non-married mothers), the sacred aura of marriage progressively vanished being replaced by parenthood norms that differ from the normative context previously associated with the traditional family based on a marital union and endorsed by law in the Napoleonic Legal Code. At the same time, the term 'démariage' raises another policy issue, namely how to handle diversity in the biographical trajectories linking children to parents. However, despite changing behaviour and value systems, marriage is still seen as the foundation of the 'true family' (Théry 2013), bringing together the three components of filiation – biological, social/educative and legal – in which both parents act as creator, carer and protector of the child in a symbolic parenthood system. This model

<sup>6</sup>Among the 9% of children living in a reconstituted family, half live with their mother and a stepfather whereas 15% live with their father and a step-mother. Slightly more than one third alternate between their two parents who both have children from the new union.

of family life, which served as a reference for both individual behaviour and policy, is no longer relevant today because of the growing number of blended families, adoption and assisted reproduction with a donor (Théry 2013). A multi-parenthood model is progressively replacing the "patrimonial" model of filiation. As a consequence, if parenthood is becoming multidimensional and diverse in its definition and practices, then it needs to be the central focus of policies.

#### *Challenging Social Policy: Lone Parenthood as a Social Risk*

In general, lone parents are more likely to be at risk of poverty and social exclusion than two-parent families, whether in access to labour markets or to an adequate income.

Whereas lone parents are heterogeneous with respect to their family trajectories, demographic characteristics, educational levels and maintenance arrangements, they share some common features. First, they generally have an income per consumption unit lower than one-earner couples (25% lower in France) and dual-earner couples with children (54%). Their poverty rate before social transfers is higher than for other households: 57% compared to 20% for French households in general, and 22% for couples with children. Social support helps to offset poverty rates: in France, the poverty rate7 for lone parents falls to 32% after social transfers, compared to 14% for total population and 13% for couples with children. Nevertheless, one lone-parent household in five depends on the minimum income (Burricand et al. 2012).

The high risk of poverty among lone parents is addressed by social policy in all European countries, with variations in incentives to participate in the labour force (Eydoux and Letablier 2009). Although some countries support lone mothers by helping with childcare, as in Ireland, others such as Sweden, and to some extent France, encourage lone mothers to return to work. In general, the conflict between working and mothering is greater for lone mothers, often due to difficulties in access to adequate and affordable childcare provision. A number of reasons concur to exclude lone mothers from full participation in the labour market, therefore making them dependent on welfare support. Only half of lone mothers have a full-time job in France, although they are the main provider of family income. The low earning power of many lone parents explains the deterioration in their living conditions. In addition, lone parents are more likely to live in rented accommodation than other households, and are less likely to own a car than other couples with children.

A major challenge for social policy is to reduce the poverty risk of lone parents by providing social transfers and/or reducing taxes. Both measures are used in France. The lone parents benefit was introduced in the mid-1970s to support lone parents during transitions. The benefit was temporary and means-tested. In contrast with the British lone parent benefit, which was provided until the child reached the

<sup>7</sup>The poverty threshold is defined relative to 60% of the median income.

age of 16 (Finn 2011), the French allowance was expected to cease when the child turned three and began attending pre-school, thereby allowing the mother to return to work. Lone-parent benefit was merged with minimum income in 2011, which means that eligibility is no longer linked to lone parenthood but to poverty. As a consequence, half of minimum income recipients are lone parents, and only 16% of these lone parents have a job.

#### *The Relative Failure of Activation Policies*

In line with the recommendations of international organizations such as the OECD, UNICEF and the European Union, policies aimed at reducing lone parents' poverty risk combine two sets of measures: providing social assistance and encouraging to work. Activation policies have targeted lone parents, notably lone mothers, with the aim of reducing their dependency on the state. Welfare-to-work policies, or workfare policies, have been developed in the 1990s onwards in order to raise the employment rates of lone mothers, thereby increasing their economic security (Letablier et al. 2011).

Lone parents' participation in the labour force varies between and within countries, depending on family trajectories, level of education and gender. In France, 23% of lone mothers have a high level of education, compared to 30% for partnered mothers, but separated or divorced lone mothers have a level of education similar to partnered mothers. Consequently, their employment rate is generally higher than for lone mothers who never lived in couple. Never-partnered lone mothers are also younger and have a lower education level. They include an increasing number of immigrant women. Overall, the activity rates of lone mothers are higher than for partnered mothers, but they are more often unemployed. When employed, they work part time less often than partnered mothers (26% compared to 34%). The proportion is lower in professional occupations (16%) but much higher for low-qualified lone mothers (39%) who are likely to be working part time because they have been unable to find a full-time job. Part-time work is generally associated with low wages, irregular income and poor quality jobs, for example in health and care services, especially in care for older people and housework. Lone fathers seem to fare better since 75% have a full-time job, though they are more likely to be unemployed than fathers living in couple.

Comparison of lone mothers' employment in Germany, France, the UK and Sweden shows that the proportion of economically inactive lone parents has declined almost everywhere, but lone parents are more likely to be unemployed than partnered mothers, the largest gap being in the UK and the smallest in Germany (Jaehrling et al. 2015). The success of welfare-to-work policy depends largely on the effectiveness of policy implementation, context (familialist or maternalist states) and on the types of jobs available. Part-time work reflects country-specific labour market patterns: in Germany and the UK, despite the high incidence of part-time work in general and of short part-time in particular, lone parents are less often in mini-jobs than women in couples; in France and Sweden, where part-time work is less common, part-time rates are also lower for lone mothers, particularly in Sweden. The specificity of France lies in the gap between high and low-qualified lone mothers and the incidence of part-time work for low qualified women when they would prefer to have full-time jobs (Eydoux and Letablier 2009). Qualitative research highlights the effect of low qualification on lone mothers' poverty risks also the effect of working conditions of the jobs they are offered: jobs are often part-time, but they are also often temporary (short-term contracts) with irregular working hours that are difficult to combine with childcare needs (Avenel 2008). Poor lone mothers who do gain access to a job are often found to be those who can rely on some form of family support, whereas those who remain unemployed are often immigrant women or women with no family or social support.

In Sweden, the labour force participation rate of lone mothers has not differed from that of mothers living in couples since the 1970s, Full-time participation is high whatever the mother's living arrangements. Mothers in couples are, however, more likely to work part-time than lone mothers (Hobson et al. 2011). Finally, the success of activation policies depends both on measures to make work pay and on services provided to lone mothers to reduce the work–family conflict.

In Italy, as in other Southern European countries where social benefits are limited, lone mothers have no other choice than to be economically active. The employment rate of lone mothers, as also the unemployment rate, is notably higher than for partnered mothers. Less than 20% of lone mothers do not participate in the labour force compared to 47% for married mothers. Similarly, divorced or separated mothers are more often in the labour force than partnered mothers. The relatively high employment rate of lone mothers in Italy goes together with the high level of education of divorced or separated mothers. A similar profile for lone mothers is found in Spain.

#### **Conclusion**

A key aim in this chapter was to explore the changing patterns of lone parenthood in European societies and to reflect on the ensuing challenges for research and policy.

A first issue was to understand how lone parenthood is defined and to what extent existing definitions provide precise, coherent and overarching concepts. Analysis reveals that the most common definitions of lone parenthood tend to oversimplify the diversity and complexity of family life as a lone parent. In some cases, the definition is too narrow, assuming for example that a single lone mother is always unpartnered and un-supported. In other cases, it is too wide, incorporating into the same concept lone parents living on their own and those living in a complex family household. In other cases, still, the definition fails to take into account an in-depth analysis of new forms of co-parenting, such as the situation of separated/divorced parents who share parental time and responsibilities with a non-residential parent. More research seems to be needed: on the complex and diverse pathways into and out of lone parenthood; on the personal and family configurations of lone parents, in order to capture their support networks and interdependencies; on the diverse meanings of lone parenting, based on the subjective choices and lived experiences of both mothers and fathers; on the impact of new fathering practices on the changing patterns of lone parenthood; on labour market segregation and how it influences gender inequalities in lone parenting. Research would therefore benefit from studies combining a variety of approaches: a qualitative as well as a quantitative approach, in order to capture the new meanings and negotiations of lone parenting from the perspective of parents themselves; a family network approach together with a demographic approach, in order to move beyond the focus on the nuclear family and on household structure at a specific point in time; a life course perspective as well as a gender perspective, in order to highlight both the different pathways and the gendered disadvantages of lone parenthood.

A second issue was to understand old and new challenges for social and family policies in this field. Poverty and social exclusion continue to be a major challenge, since there are lone parents who suffer from a combination of several types of disadvantages: gender (mothers), younger children, lower educational attainment levels, immigrant background, unstable or part-time jobs in the labour market. Analysis also reveals a new trend in policy-making: a move away from a family policy perspective focusing on the vulnerability of lone parent families in general and a shift towards a social policy perspective seeking to provide social benefits for disadvantaged persons and families, with little concern for specific family type. The consequences and effects of such a trend (e.g. in France) have yet to be analyzed. They are likely however to increase stigmatization of young low educated lone mothers, those who find it more difficult, in all European countries, to move from welfare to work and to be residentially and economically independent in order to bring up a dependent child on their own. Raising awareness and debate on this issue and how it may reproduce stigmatization is an important challenge for social policy research.

Policy responses to changing patterns of parenthood are a last major challenge. The higher the number of lone parents, the higher their diversity in European societies. Lone parenthood is nowadays one among many contemporary family forms. Whereas these living arrangements have increased markedly over the past 50 years, albeit to varying degrees across countries, their social acceptability has also changed considerably. For children, the complexity of living arrangements means that, instead of being assigned throughout their childhood to a lone parent, they often have several parents each performing different roles. Sociologists and policy makers have attempted to redefine lone parenthood without providing a satisfactory answer so far as to how to take proper account of the many changes in living arrangements from a life-course perspective.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **References**


**Marie-Thérèse Letablier** Senior Researcher Emeritus, Centre national de la recherche scientifique – CNRS – and Centre d'Economie de la Sorbonne – CES – Université Paris1 Panthéon-Sorbonne. Her main research interests are in family policies, family welfare, fertility, gender equality and work-life balance, in a European comparative perspective.

**Karin Wall** Research Professor at the Institute for Social Sciences at the University of Lisbon (ICS-UL). Her main areas of expertise are comparative social policy and sociology of families and gender. Her research interests include family policies, family forms and interactions, gender equality and work-family balance, lone parent and migrant families, family trajectories and the life course.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Chapter 3 A Media Discourse Analysis of Lone Parents in the UK: Investigating the Stereotype**

**Emma Salter**

#### **Introduction**

Lone parents are a heterogeneous social grouping (May 2006); lone parenthood can have multiple causes, whether divorce or separation, widowhood, lone parenthood by choice or by necessity (Giddens 2006). While the majority of lone parents in the United Kingdom are women, men currently comprise 8% of lone parent families with dependent children (Office for National Statistics (ONS) 2013). The lone parent population increased from 840,000 to 1.6 million between 1979 and 1997 (The Times 1993a; ONS 2013), but has since grown at a slower rate with recent figures putting the number of lone parent families with dependent children at two million in 2014 (Gingerbread n.d.; ONS 2014).

Historically, women who had children 'out of wedlock', together with their 'illegitimate' children, have either hidden within family myths of daughters growing up as sisters or were sent away to spare the family their supposed shame, with lowerclass children often adopted by middle-class married couples (Thane and Evans 2012). In more recent times, with greater variety in family formation and more acceptance of cohabitation as an alternative to marriage, with or without children, such views seem outdated and out of place. However, lone parents remain highly stigmatised in the United Kingdom (Hinton-Smith 2015) to the extent that lone parents themselves make a concerted effort to self-identify as 'good mothers' to distance themselves from the 'bad' sort (Phoenix 1991) while one of the missions of Gingerbread, the charity supporting lone parents in the UK is to "dispel the myths and labels" around lone parenthood (Gingerbread n.d.).

All too often the term 'lone parents' is used by politicians and in the media, without defining which lone parents are meant and thereby classifying all lone parents as problematic. For example, lone parents have been labelled "one of our

E. Salter (\*)

School of Education and Social Work, University of Sussex, Falmer, UK e-mail: e.j.salter@sussex.ac.uk

<sup>©</sup> The Author(s) 2018 55

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_3

greatest social problems" (Tom Sackville, then junior Health minister, quoted in White 1993) and identified as the targets of a Conservative backlash (Moore 2013b). It is unlikely in either case that such a statement was intended to encompass all types of lone parents, of all backgrounds, ethnicity, age and gender, yet this lack of distinction appears to be commonplace in ministerial pronouncements and newspaper articles. This chapter presents the results of an intersectional analysis of newspaper articles from two years spanning two decades in order to uncover which identity factors are most commonly assigned to the lone parents discussed within these media sources and by politicians. First, however, intersectionality theory and its utility for this purpose are outlined.

#### **Intersectionality**

The term intersectionality first gained recognition in Kimberlé Crenshaw's 1991 article in the Stanford Law Review, however, the concept was in existence in a number of forms before this as a response by the Black Feminist Movement in the 1960s to the white, middle-class domination of feminist theory. The criticism was, rightly, that while (predominantly White, middle-class) feminists were fighting for the rights of women; they were not representing the needs and issues of all women, but their own concerns, rather than for example, those of women from the working class or from other ethnic groups. Crenshaw, a lawyer, wrote that while sex discrimination law was providing a legal framework to safeguard the rights of White women and race discrimination law was protecting Black men there was no legal protection where these identities overlapped, that is in relation to Black women (Crenshaw 1991). Her analogy for intersectionality was as roads of discrimination (e.g. race, gender, sexuality) which met at intersections or crossroads, emphasising the multiple identities that intersect. Like others before her, for example Deborah King's 'multiple jeopardy' (King 1988) or Patricia Hill Collins' 'matrix of domination' (Hill Collins 1990), the emphasis was on sites of multiple oppression and the negation of ideas of additive disadvantage, rather they argued, these disadvantages were multiplicative.

While initially the social markers of gender, race and class were the main foci for intersectional scholars, subsequent development of the theory of intersectionality has broadened the categories to include others such as sexuality, age or ability. Each combination of these identifiers positions us on what Yuval-Davis terms a 'power axis of difference' (Yuval-Davis 2006). As Yuval-Davis points out, official statistics tend to assign people to one of these power axes, but each individual is in actuality situated on multiple axes, which, depending on the context, assign more or less power to the individual. For example, in the UK, being a member of the privileged categories of male, White, heterosexual and middle-class brings certain advantages, while if even one of these categories changed to a more minority alternative, whether in terms of gender, class, sexuality or ethnicity, this would change the cumulative advantage. In addition, this (dis)advantage would differ according to which social location was changed and how. Layers of oppression and privilege exist in most if not all societies, which enable a White middle-class woman to enjoy advantages in the UK, which could be lessened, or non-existent, if she were either White and working-class or Black and middle-class, but even these power dynamics are context dependent; for example, a White middle-class woman might feel disadvantaged if she were in a context where she was now in the minority.

Intersectionality is important in the discussion of perceptions of lone parenthood as it allows us to investigate the intersections of these multiple identities which in turn reveal the power dynamics of the dominant discourse. Modern Britain is not an equal society on many levels (Hills et al. 2009), with those who are outside the norm often constructed as deviant (Wilson and Huntington 2006). As Garner has noted, these norms are "usually class-based, gender-biased and ageist" (Garner 2007; p. 6). Intersectionality as an analytical approach enables the identification of the multiple and interconnected social positions of lone parents and, consequently, where they are located in modern British society.

#### **Data and Methodology**

The corpus for analysis consists of articles from two sources from two years, 1993 and 2013 which frame two decades in which the numbers of lone parents have stabilised in the UK. The sources are two national broadsheet newspapers, The Times and The Guardian, which represent respectively the two ends of the political spectrum in the UK. The Times is a right-wing newspaper with a large Conservative readership (The Independent 2015), while The Guardian is the most left-wing of the broadsheets (BBC 2009). Although their distribution figures are smaller than those of the tabloid newspapers, as broadsheets they were less likely to present a sensationalised coverage of lone parenthood.

The years selected for analysis, two decades apart, were chosen for a number of reasons. Firstly, 1993 has been described as the 'year of the lone parent', a key year in the 'moral panic' surrounding lone parents (Mann and Roseneil 1994). A combination of factors not only put lone parent families in the spotlight, but blamed them for both the burden on the public purse and the perceived upsurge in juvenile crime. The increase in the numbers of lone parent families was seen as key to the burgeoning social security budget, with the result that the Child Support Agency was created to recoup money from absent fathers. In addition, concerns about juvenile crime and poor parenting were prevalent in the light of the murder of the two year old James Bulger in February 1993 by two ten year olds (one of whom was from a lone parent family, the other whose parents were separated (Faux 1993)).

As can be seen from Fig. 3.1, there was an upsurge in media interest in lone parents in 1997/98, particularly by The Guardian, coinciding with the arrival of the Blair government and a number of policy initiatives focussed on lone parents. However, an analysis of articles from this period, only five years on, would not have had the benefit of taking a longer view on how lone parents were portrayed.

**Fig. 3.1** Number of articles with single parent\* mentioned anywhere in the text in The Times and The Guardian (1990–2014)

Media interest declined soon after in the run up to the millennium, before increasing again a decade later, with the highest number of mentions occurring in 2012 and 2013. The introduction of Universal Credit in 2013, a streamlining of the complex benefits system, and of a benefit cap in the same year (Department for Work and Pensions 2015), both of which have been shown to disproportionately penalise some members of the population, including lone parent families (Whitworth 2013), contributed once more to bringing lone parent families to the media's attention. Since the numbers of lone parents had stabilised in the intervening years, it might be hypothesised that this would result in a more accepting attitude towards lone parenthood than two decades previously. These factors combined to identify 2013 as an appropriate time point at which to examine whether lone parents were discussed in similar terms as 20 years previously.

To identify the articles for analysis, Nexis® was used to search the chosen sources for articles relating to lone parenthood in the selected years.1 Specifically, all articles containing major mentions (that is which appear "in the headline, lead paragraph or indexing" (Lexis Nexis 2015)) of the following terms: single parent, one parent, lone parent, single mother, single father, single mum and single dad were identified for analysis; resulting in 1081 articles (see Table 3.1). Subsequently, duplicates, primarily a result of different editions of the newspaper being archived, and all non-news articles, for example, those discussing books, films or television programmes were then deleted. This caused the total for the 2013 corpus in particular to drop considerably, indicating that even if lone parents were not making the headlines to the same extent 20 years later, they remained a subject of considerable interest as the subject of novels and on-screen dramas. Finally, only articles which

<sup>1</sup> In British English, unlike in the case of some European countries, the two terms 'lone parent' and 'single parent' are interchangeable; both signify a parent who is the sole resident carer of their child(ren), no matter the cause and both allow for the presence of a non-resident father in the child(ren)'s life.


discussed lone parenthood in the UK were retained, resulting in a final corpus for analysis of 631 articles, across sources and years**.**

As can be seen from Table 3.1, there were considerably more articles that fulfilled the search criteria from The Guardian in 1993 than its right-wing counterpart, both initially and in the final analytical corpus, whereas the totals are more comparable in 2013. It remains to be seen whether the discourse on lone parenthood is similar in the two sources and whether there is continuity over time.

The corpus was hand-coded using the NVivo software, with all identity factors that were used in connection with lone parents coded. Following the key identity factors embedded in intersectionality theory, the analysis initially focussed on locating mentions of demographics such as gender, ethnicity and class in relation to lone parents. In the process of coding these factors, it became clear that other social locations such as age, sexuality, income and the causes of lone parenthood were additionally used within the corpus to identify a certain type of lone parent, so these were added to the analysis.

The quantitative results of the analysis are set out in Table 3.A1, detailing the number of references found for each element of identity, disaggregated by newspaper and year. The following section examines each identity factor in turn. Each subsection presents the results of an analysis of the 1993 corpus for both sources, followed by the findings of the 2013 analysis, noting any similarities and differences between the two sources and time periods.

#### **Identity Factors**

#### *Gender*

When lone parenthood is mentioned, inevitably women are at the forefront of the discussion, since they represent the overwhelming majority of lone parents, while men currently comprise only 8% of the UK's lone parents with dependent children (Gingerbread n.d.). It was anticipated therefore that an analysis of the corpus would reveal a significant number of references to lone mothers, single mothers and single

mums, and proportionally fewer mentions of single fathers, lone fathers and single dads. Since the gender breakdown of lone parents in 1993 is reported variously within the corpus as 90% and 95% female, it is understandable that there would be a greater focus on lone mothers than on their male counterparts. This was indeed the case; however, the references to lone fathers are even fewer than expected, forming fewer than 2% of all references to lone parents.

Despite their scarcity, these articles provide some noteworthy details about lone fatherhood in the Britain of the early 1990s. One reveals the little-known statistic that, in 1993, Britain had one of the highest incidences of single father families in Europe (Carvel 1993), a fact that is all the more interesting for not being referred to anywhere else in the corpus, either by journalists or politicians. Another article reports survey results in which a third of British women felt that a lone mother could bring up their child as well as two parents, while fewer British women and men agreed that a lone father could do likewise (McKie 1993). While lone fathers were in the minority in the UK's population of lone parents, it is notable that their growth in numbers was not seen to be as of great a concern as the increased numbers of lone mothers, particularly if they were perceived as less able than women in nurturing their children. Unlike lone mothers, there are only a few examples of articles where elision occurs between lone fathers and lone parents; references to lone fathers tended to either to be the focus of an article or in response to such a piece, rarely were they explicitly included in wider discussions of lone parenthood.

The majority of references to lone parents, were made in gender-neutral terms, with references such as lone parent(s), single parent(s) and one parent occurring more frequently than their feminine counterparts in both sources and time periods. While this is explained in part by references to lone parent benefit and other genderneutral policy terms, the practice also extends to discussions of the lives of lone parents by both journalists and politicians. The majority of female and genderneutral references occur separately from one another, that is, articles use only one or other of these to refer to lone parents throughout. Nevertheless, over 10% of articles use references to lone parenthood and lone motherhood interchangeably, or use lone parent in reference to a mother. Some of these are for reasons of clarification, for example, "the number of lone parents, 90 per cent women" (Brindle 1993a), yet other articles and quotes from politicians slip seamlessly from gender-neutral to feminised depictions of lone parents for example,

The controversy's roots go back to government attacks on single mothers during the Conservative Party conference back in October. A parade of ministers lined up lone parents as the villains of welfare spending (M. Phillips 1993).

and,

Margaret Thatcher said it would "give the lone parent back her morale and her confidence" (The Times 1993b).

The same phenomenon occurs in headlines in the 1993 corpus of The Guardian; on three occasions the gendered nature of lone parent descriptors are changed between the headlines and subheadings, thereby revealing that it is in fact issues around single motherhood that will be the article's primary focus (Hetherington 1993; Griffin and Younge 1993; Weston 1993).

In terms of shifts in the discourse over time, two decades later, there are noticeable differences in how gender is referenced. Lone fathers make up a larger proportion of the articles than in 1993, although still not commensurate with their proportion in the population. Lone mothers are discussed in nearly two-thirds of articles in The Guardian, while The Times uses gender-neutral and feminine referencing fairly equally. There are very few elisions between lone parents and lone mothers in the 2013 corpus, with their usage almost completely distinct.

Overall, the quantity of direct references to female lone parents in the corpus, coupled with the number of indirect references where gender-neutral terms become female through clarification or juxtaposition, indicates that when these media sources discuss lone parenthood, their focus is lone mothers rather than lone fathers.

#### *Ethnicity*

In the UK, the use of ethnic descriptions tends only to be used to identify people who differ from the majority ethnic group of White British. Whiteness is an assumed, unmarked category, by nature of its normativity (Garner 2007), so we would not expect White lone parents to be identified as such; rather we would expect ethnic markers to be present for ethnicities other than White and whiteness to be marked by an absence of ethnic descriptors.

As can be seen from Table 3.A1, the identification of lone parents in ethnic terms is indeed largely absent from the corpus. Whiteness is referenced nine times in the whole corpus, compared to 13 references to African-Caribbean or Afro-Caribbean single parents and 12 to the more generic Black category. There are no discussions of ethnicity relating to lone parenthood in The Times in 2013.

With very few exceptions, ethnic identities are mentioned either when they are the focus of an article or when comparisons are being made about the proportions of lone parenthood between ethnic groups. African-Caribbean families were most likely to be presented as the comparative category of interest, primarily because "in comparison with a 14 per cent figure in the general population, single mothers accounted for 51 per cent of all Afro-Caribbean births" (M. Phillips, 1993). That 'Afro-Caribbean births' are compared with births in the 'general population' is indicative of the normativity of White ethnicity, with other ethnicities situated outside. The Times coverage in 1993 refers to more ethnic groups than The Guardian: The Guardian refers to 'Asian mothers' once in the context of comparing their lower proportion of lone motherhood with the White British population, while The Times refers several times to Asian populations, and additionally to Indian and Pakistani families.

In 2013, The Times does not discuss lone parenthood in ethnic terms, but The Guardian's references to ethnicity are for the same reason as in 1993; comparisons are made between the incidence of lone parenthood in the White population and other ethnicities, with a focus (again) on African-Caribbeans, who are "twice as likely… to grow up in a single-parent household" (Corner and Normanton 2013).

Aside from referring to ethnicity in the context of comparative statistics, articles mentioning ethnicity in a substantive way in relation to lone parenthood all focus on the issues the Black community face. Journalists debate the challenges of changing cultural habits in the Black community such as men with several 'babymothers', with whom they have children but whom they may or may not support (Ford 1993) and question the stereotype of whether Black lone mothers are the passive victims of such Black male behaviour (M. Phillips 1993b). Where ethnicity and lone parenthood, therefore, appeared together in a few articles in the corpus, the norms and behaviours of an ethnic group (in all cases the Black, or more specifically, the African-Caribbean population) were the main subject of the article, with other ethnic groups only mentioned for comparative purposes.

Whiteness is referenced only three times in relation to lone parenthood outside of those articles comparing the proportion of lone parents in different ethnic populations. On two occasions the article is penned by the single parent in question who self-identifies on ethnic lines, yet in both articles whiteness is not mentioned incidentally, but in order to make a point. In the first, the writer is a White middle-class male who on becoming a single parent, suddenly finds himself "to be a minority within a minority" (Bovill 1993), the irony being of course that this is not the usual situation for someone with his particular intersections of class, gender and ethnicity. In the second, the author of the piece is a White single parent to two mixed race girls, and the subject of the piece is about racial identity and dealing with the cultural diversity of her family. In the third article, the situations of two lone parents are discussed: one who is White with "two pretty blonde daughters" and the other who is Jamaican (Norman 1993). So, whiteness is explicitly stated only on those occasions where the fact of being white matters to the piece as either a counterpoint to a different ethnicity or to emphasise the irony of a situation. Otherwise, whiteness is unmarked, for example, in an article about three lone parent families appealing the introduction of the benefits cap. Two of the mothers are identified as being respectively Roma and Orthodox Jew, while the third is not assigned any ethnic identity; the assumption being that she is White British.

In sum, the majority of references to lone parents in the corpus are not ethnically identified, except where a point is either being stressed about comparative numbers of lone parent families in Black or Asian populations or where the cultural stereotypes of an ethnicity other than White are under discussion, though here again the focus is on African-Caribbean families. Whiteness is asserted when it is necessary as a juxtaposition to other ethnicities, or as an ironic aside, otherwise it is absent from the debate. The logical conclusion therefore is that, in both years, and both sources, the unmarked category of whiteness is the assumed normative ethnic identity of the lone parents being discussed, so remains unstated.

#### *Class*

An examination of the corpus for explicit references to "class" reveals that direct allusions to class are, with a few exceptions, absent. As can be seen from Table 3.A1, there were 14 explicit references to class in the 1993 corpus, evenly distributed between the two newspapers, and one in the 2013 corpus. Of these 14, only three refer to middle class lone parents, the rest refer to lone parents in the lower classes or working class. A couple of Guardian journalists write, for example, that "most lone parents tend to belong to social economic groups 4 and 5" (McGlone 1993) or "a working class estate with high unemployment and a high proportion of single parents" (Katz 1993). Additionally, there are a couple of indirect references to the classed nature of lone parenthood, for example, the following from a letter in relation to fathers rallying against the Child Support Agency: "The mighty middle classes flex their muscles and the Government considers a U-turn. A pity single mothers don't have that sort of clout" (Russell 1993) which, indirectly implies that the single mothers are not middle class since they are do not have the influence of the "mighty middle classes". In The Times' coverage from the same year, a slightly different emphasis emerged. Discussion of class is still largely absent, comprising only six references in total, but four of these six use the term 'underclass' applied specifically to lone mothers. The term was introduced to the UK by the American social theorist Charles Murray in The Sunday Times, to refer to those people who exist at the lowest level of society, so it is perhaps anticipated that the term would be taken up by its sister publication in this context. The other two references are to class in general and to "young working-class women living on low incomes" who are identified as comprising the largest proportion of lone parents in contrast with a "small percentage…accounted for by widows and middle-class divorcees" (Dynes 1993).

Similarly, the 2013 articles from The Guardian include three direct references to class and lone parenthood, and in each of these, lone parents are seen as part of the working-class. One explanation for the lack of explicit references to class, is that, "Class hatred has been siphoned off on to chavs, scroungers, benefit fraudsters, single mothers, all the new untouchables" (Moore 2013c), indicating that it is perhaps not class that is relevant now, but different categories of people in society, although associating single mothers in this way with scroungers and benefit fraudsters places them by implication in the same economic category, that of being dependent on the state. There are no direct references to class and lone parenthood in the 2013 corpus for The Times.

While class is not widely referenced, it appears that when it is mentioned, that the lone parents are more often associated with the working-class, lower classes or underclass. Although explicit references to class were few and far between, it became evident while coding the corpus that proxies for class, such as receipt of benefits, housing tenure or income levels were more prevalent in the discussion of lone parenthood. It is to this we now turn.

#### *Economic Factors*

It is unsurprising that social security received a great deal of media attention in 1993, since the main thrust of Conservative policy at that time was the reduction of the social security budget. It is unfortunately also the case that relationship breakdown often results in a large and sudden reduction in income, particularly for women (Mortelmans and Jansen 2010) leading to an increased reliance on public funds and resources, even if temporarily. These two factors mean that in 1993 there was a political, and as a result, media focus on lone parents', specifically lone mothers', dependence on state support.

As can be seen from Table 3.A1, there were a large number of references in 1993 to benefits and council housing in relation to lone parenthood, not least due to the proselytizing of Conservative ministers at their Party Conference that these were motivating factors for the increased numbers of lone parents in the UK. Articles in both sources not only reported these speeches, but continued to do so even once it was revealed that a Cabinet paper with evidence that such associations were without foundation had been circulated weeks before the conference (Brindle 1993a). Meanwhile, every repetition of such unfounded statements, even if refuted, only served to reinforce this inaccurate stereotype.

Some journalists took a different approach, with a focus on the issues of benefit dependency and lone parenthood and the need for greater Government spending on childcare in order to provide "a pathway out of the poverty trap" (Taylor 1993) and enable lone parents to (re)enter the labour market. However, there are very few articles in the corpus, even from left-wing journalists, which criticize the raft of policy proposals made by the Conservative Government which were aimed specifically at lone parents: the phasing out of lone parent benefits, cutting benefits to lone parents who had additional children while claiming benefit (Brindle 1993c), limiting access to council housing (The Guardian 1993b), cuts to education funding in those councils with larger numbers of lone parent families (Wainwright 1993), and finally, at the instigation of the Child Support Agency, penalties for lone mothers who refused to name the father of their child, whose benefits could be cut by a fifth for 18 months (Baxter 1993). If there is any criticism it is that the Government and the Child Support Agency were ignoring the impact of such policies on the children of these families, but even these critiques come from external sources, not Guardian journalists. For example, in an article in the Money section the Child Poverty Action Group are quoted as saying that the Child Support Agency had "one rule for the rich, another for the poor" since penalties such as a reduction in benefits for not naming the absent father would (and presumably could) only be imposed on lone mothers who were benefit claimants (Hughes 1993).

This discussion of differentiation by wealth is to be found in two articles in the 2013 corpus, which reveal that little has changed in this regard in the intervening years. In one, Suzanne Moore writes on the Government's 'moralizing' about lone parenthood, remarking that in the view of the Conservative government, lone parenthood is tolerated if you can support yourself, but if you are on benefits, then you are a "subspecies in need of help" (Moore 2013a). In the other, Zoe Williams reinforces this opinion, accusing the Conservatives of only taking issue with poor lone parent families, while disregarding the behavior of the rich (Williams 2013).

Another theme which emerged from both papers in the 1993 corpus was a propensity to link lone parenthood with deprivation and poverty. Journalists reported statistics, for example, "seventy-five per cent of single parent families live in poverty" (Moore 1993), but also used turns of phrase which implied that the proportion of lone parents added to the deprivation of an area. For example, "The Sixties' estate bears all the hallmarks of inner-city deprivation – 90 per cent of this year's nursery intake is from young, single-parent families; 77 per cent of residents receive benefit" (Thomson 1993). A further example goes so far as to place lone parenthood on an equal footing with 'economic deprivation' and 'bad housing', stating that, "we can argue all around the houses about the relative effects of these three factors on the behavior and development of young people" (A. Phillips 1993). Other articles reported the inclusion of lone parenthood as one of six indicators in the Government's social deprivation index and one of three "traditional needs indicators" for apportioning education funding to councils (Wainwright 1993). It appears that lone parenthood and poverty had become synonymous.

In 2013 although the number of articles referring to economic factors in relation to lone parenthood are far fewer, these connections between poverty and lone parenthood remain: "the poorest households – such as single parent households with children" (Butler 2013). In fact, despite fewer references in the later corpus, children of lone parent families were "twice as likely to live in poverty" in 2013 than those from two-parent families (Paton 2013). It would seem that the intervening years of social policy from successive Governments had worsened rather than improved the conditions in which lone parents and their children live. In summary, while there are some references to class in relation to lone parenthood, certainly in the 1993 corpus, both newspapers are more concerned with lone parents' lack of income and consequent dependence on the state.

#### *Age*

Age is not often the first aspect of identity that presents itself when considering the different social locations of individuals, but in relation to women and childbearing, it is vitally important. The age of a woman when they have their first child is under constant scrutiny: too young and it is considered a public health problem (Lawlor and Shaw 2002), too old and it could put mothers and their babies at risk medically, while socially, it can be viewed as selfish (Hadfield et al. 2007). That age represents an important issue within the discourse of lone parenthood is supported by the choice of Government statistics contained on the Gingerbread website. The first set of facts relates to the numbers of lone parents in Britain, but the next specifically targets perceptions around age and lone parenthood, with the statements that "less than 2 per cent of single parents are teenagers" and that the "median age of

single parents is 38.1" (Gingerbread n.d.). The small percentage of teenage single parents is an important reminder to those who believe that teenage sex is rife in the UK; in a recent poll, the British public estimated the percentage of teenage pregnancy as 25 times greater than official figures (Ipsos-Mori 2013). Likewise, the stated average age of lone parents points to a very different life stage from the teenage years. Both facts are aimed at rebutting the misconception that lone parenthood equals teenage parenthood.

In the realms of lone parenthood, therefore, youth is pinpointed as a decisive issue. In terms of the assignation of demographic characteristics within the corpus, age, or more specifically, youth, is the most used identifier in terms of lone parenthood, second only to gender. As can be seen in Table 3.A1, in the 1993 corpus, youth is mentioned 60 times in The Times and 81 times in The Guardian. Although there is one mention within the 1993 corpus of a significant drop in the numbers of teenage lone parents by 1993, the majority of articles reflect an emphasis on the youth of lone parents in the UK. The repetition of the then Social Security Secretary Peter Lilley's nonsensical parody which stated that teenage girls were economicallymotivated to get pregnant deliberately certainly increased the attention paid to younger mothers. The Guardian quoted a number of research reports quashing that notion, but still felt the need to reiterate the established trope that they were contradicting, giving it more column inches in the process.

Not only are these women young, but they are also contradictorily portrayed both as becoming pregnant on purpose in order to receive benefits and preferential treatment for council housing, and pregnant by accident as in this quote from Sir George Young, Housing Minister:

How do we explain to the young couple who want to wait for a home before they start a family that they cannot be rehoused ahead of the unmarried teenager expecting her first, probably unplanned, child? (Young quoted in Brindle 1993b)

As youth is subjective, it might be argued that the numerous references to young mothers (there is no reference in the 1993 corpus to young fathers) are not necessarily a fixation with the teenage years. However, an examination of the corpus reveals that, apart from one statistic denoting the proportion of lone mothers under the age of 30, age was otherwise referred to via a number of descriptors, all of which positioned young parents in the teenage years. Examples such as "under 20", "before they are old enough to vote" were found, as well as less arbitrary descriptors such as 'gymslip', 'schoolgirl' and 'teenage'. One article even differentiated between those who conceived and those who gave birth as teenagers: "research suggests that 25 per cent started as teenagers and 33 per cent first became pregnant when under 20."(The Guardian 1993a).

In 2013, there are only a handful of mentions of age, yet again, as in 1993, aside from a couple of features about older lone parents – all of whom are successful career women – it is primarily the younger members of the lone parent population who are identified in age terms. The references appear in relation to two main issues, firstly, budget cuts to hostels for the under-25s in a London borough, with the potential for young parents to be moved hundreds of miles from home. Secondly, and somewhat ironically considering the 1993 rhetoric on young mothers and council housing, a report on proposals by a group of Conservative MPs which threatens to deny social housing to "Britain's youngest single mothers …as part of a new drive to reduce teen pregnancy". Under such proposals these young mothers would be forced to live either with their parents, or in the hostel accommodation currently having its budget cut or risk having their benefits removed. While the age of lone parents in 2013 is not the key concern in these newspapers that it was in 1993, there are signs that teenage mothers have remained in the crosshairs of policy-makers, 20 years on. Finally, the issue of age and lone parenthood rarely refers to fathers: in the 2013 corpus, young fathers are mentioned in one article on young, black fathers. White young fathers are entirely absent from the corpus.

#### *Causes of Lone Parenthood*

The causes of lone parenthood are multiple, whether from separation, divorce, desertion, domestic abuse, choice or necessity. I was interested, therefore, to discover which of these were most debated in the selected press, by journalists and politicians, since as we have already seen, the homogeneity of the discourse can belie the multiple identities contained within.

On occasion, politicians were quoted as being aware of the heterogeneity of lone parenthood and pledged that they had differentiated between these categories, for example, this from John Redwood, the then Welsh Secretary:

I was very careful to distinguish between different types of single parenthood. I've always felt extremely sympathetic to those who are widowed, to mothers who are beaten up or abused, or to fathers and mothers who are on the wrong side of a losing relationship, often through no fault of their own. (Redwood quoted in Hetherington 1993)

Redwood's comment was in response to allegations that all lone parents were being labelled in the same way. While he expresses sympathy with those he includes above, he goes on to say that "society has a role to play in encouraging young girls to knuckle down at school, to think about a stable relationship before having babies" (Redwood quoted in Hetherington 1993) which indicates that this trope about young girls having babies outside of marriage, or even a stable relationship, is the real concern when it comes to lone parents.

His remarks are indicative of the stance of the Conservative Right at the time; lone parents as a whole were not seen as a problem, just a subsection, yet they still used generic terminology as shorthand, even if they "know…how furious it makes the divorced, widowed and deserted, struggling alone, when headlines say "Ministers attack lone parents"" (Peter Lilley, quoted in Grove 1993).

Beyond the quotes included above, the broader causes of lone parenthood are little discussed within the corpus. There are only a handful of further mentions of "deserted/abandoned", or "abused" lone parents, leaving the majority of references to causes of lone parenthood to focus instead on marital status or family transitions. For example this from The Guardian, discussing research using the National Child Development Study: "The children can be divided into four groups: those in twoparent families, or the three forms of single parent families: never married; single but divorced; single by the death of a partner" (Dean 1993). Or this from The Times: "In the past decade the number of births outside marriage has more than doubled to one in three, a rise produced by the growing number of single, divorced and separated mothers" (Dynes 1993).

In fact, the discourse on types of lone parenthood in the corpus reflects to a large extent the emphasis of Redwood's statement. Over half of the references to lone parents' marital status in the 1993 corpus defined them as 'unmarried', although surprisingly The Guardian articles included twice as many as the pieces in The Times. The unmarried mother was variously "married to the State" (a phrase from the American social theorist Charles Murray adopted by the Conservative Right), or were responsible for 'spawning' a "welfare-dependent underclass" (Baxter 1993). Separation and widowhood are mentioned but infrequently. References to divorce tend to appear less in regard to defining the cause of lone parenthood, but more often are positioned alongside lone parenthood in discussions of family trends, for example, "The combined force of single motherhood, children being born out of wedlock, divorce, remarriage and the rest" (Wicks 1993). Aside from the difference noted above in the number of references to unmarried lone parents between the two sources, a more nuanced distinction can be seen in how statistics on the circumstances of lone parenthood are reported. In 1993, in The Guardian, the latest statistics were reported as follows, "Welfare groups last night pointed out that the political concern with single mothers, who have never married, obscured the fact that about two-thirds of lone mothers are divorced, separated or widowed" (Wintour 1993), the same facts are reported in The Times with a change of emphasis, "Single parents include widows and divorcees, but the fastest growing group are the "single, never partnered", who account for more than a third of all lone parents" (The Times 1993a).

Analysis of the 2013 articles reveals very few mentions of the causes of lone parenthood, in either newspaper, whether family transitions, marital status or other causes. There are so few references even to marital status that it is hard to draw a conclusion about the discourse; it appears that the causes of lone parenthood or the marital status of lone parents is no longer of great interest in 2013, which, it is hoped is a positive step towards a more tolerant discourse. Interestingly, in neither year, in neither paper, is there any discussion on the temporary nature of lone parenthood. Statistics put the average length of time a child spends in a lone parent family as seven years (Gingerbread n.d.) yet the way it is discussed by journalists at both ends of the political spectrum lends it a greater sense of permanency.

#### *Sexuality*

It is only since 1999 that the number of dependent children living in same sex couple households has been included in official government statistics about the family and the household and even these estimates are, by their own admission, not reliable (ONS 2014). This suggests that it would be unlikely that sexuality would form a key topic in the discourse on lone parenthood. However, there are now more options available for people of different sexual orientations to become parents and they, like heterosexual couples, are not immune to break-up, so it was possible that the 2013 articles may have included some discussion on this topic.

Nonetheless, the analysis unearthed very few references to sexuality in the corpus in the context of lone parenthood. In The Guardian in 1993, the majority of references to sexuality in this context are with regard to the unequal footing of lesbians, alongside lone parents, rather than lesbian lone parents, whether in terms of access to council housing, donor insemination clinics or as the target of the Conservative Right as unfit parents (Neale 1993). Except for one reference to sexuality and lone parenthood in 2013 in The Times (Slater 2013), it is not clear whether the lesbian women approaching the clinics are in partnerships or not, but it seems that in general, there were two separate discourses in circulation, one concerning lone parents and another gay parents. In 2013, Suzanne Moore makes the point that in a society where the ideal is still heterosexual marriage then lone parents are as acceptable as gay parents (Moore 2013a). Aside from the one article about a lesbian lone mother, discussion of sexuality is absent from the discourse of lone parenthood in both years. It must be deduced that, just as with ethnicity and gender, the normative of the heterosexual is assumed in these media sources' depiction of lone parenthood.

#### **Discussion and Conclusion**

I set out to discover whether the generic category of 'lone parent', that is, the lone parent population as a whole, was the intended focus of political and media interest in the early 1990s and 2010s, or if a more nuanced picture would emerge from a detailed analysis of articles in two broadsheet newspapers at the time. I chose an intersectional lens in order to look at the intersections of lone parent identities included in the corpus. As a first stage, it was necessary to identify the prevalence of different identity factors, as discussed above, before seeing if and how these intersected.

A key finding of the analysis is that there are very few occasions where intersectional identities are applied to the lone parents under discussion. An attempt to identify instances of an intersectional identity according to traditional intersectionality lines, that is gender, ethnicity and class, returns only three results, all of which were discussed in the ethnicity section: one, the self-identified middle-class White male, and the second and third, the lone mothers, one White and one Black, living on benefits.

This is partly due to a lack of references for some social locations, for example, the absence of markers for ethnicity, class and sexuality, together with the prominent usage of gender-neutral terms for lone parents. If the search is broadened to encompass the other social locations discussed such as age, sexuality and marital status, there are a handful more, most focusing on a combination of youth, unmarried and female as identifiers, with additional allusions to benefit receipt or council housing, or youth, class and gender, for example "young working-class women living on low incomes" (Dynes 1993).

Despite the lack of intersectional identities within the corpus, we can conclude that the identity of these media sources' focus in 1993, the "year of the lone parent", can be defined in terms of gender, ethnicity, class, income, age, marital status and sexuality. She, for it is a she, is White, from the lower classes and economically reliant on the state. She is young, a teenager, who is unmarried and heterosexual. In 2013, this pen portrait is just as apposite; although the corpus of articles was far smaller than in 1993, the results of the analysis paint the same picture (excepting any discussion on marital status), 20 years on.

Despite the prolific usage of gender-neutral terminology such as lone parents, single parents and one-parent families, an intersectional approach to these sources reveal that the media portrayal is far from being an all-inclusive term which homogenizes the experiences of all lone parents. Rather, hidden beneath the generic terms, is a clear picture of the true subject of the media discourse. The benefit of an intersectional standpoint is in the multiple identities captured within the analysis, which together provide a complete picture of the subsection of lone parents who are the subject of media focus. Of these identity factors, whiteness and heterosexuality are usually seen as positions of privilege within the intersectionality discourse, yet in this instance, the other identity factors remove such privilege one by one. To be female places a lone parent at a social disadvantage, to be young is to be disempowered, to be working-class (or worse, part of the underclass) is to be in a socially inferior position, to be unmarried, is to be without the economic support of a partner and therefore stigmatised by reliance on the State. Together these identity factors result in a multiply disadvantaged social positioning. The analytical framework reveals that the specific identity of lone parents, beleaguered by politicians and the media, is representative of some of the most vulnerable people in UK society, both in 1993 and 2013. As one features writer said, "A teenage mum pushing a pram round a run-down estate is an easy target" (Holmes 1993).

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **Appendix**


**Table 3.A1** Number of times identity factors were coded in corpus, by year and source

a Variously defined as lower orders, Socio-Economic Group 4–5 and lowest social grouping

#### **References**


Phillips, M. (1993b, December 18). Another type of family value that's faring well. The Guardian. Phoenix, A. (1991). *Young mothers?* Cambridge: Polity Press.

Russell, L. (1993, November 5). Letter: Child's play for middle-class muscle. *The Guardian*.

Slater, L. (2013, November 30). Baking with my daughter. *The Times*.


**Emma Salter** is a PhD student at the University of Sussex, investigating the educational and social outcomes of the children of lone parents. She has worked as a researcher at the University of Sussex and the Institute of Education, University of London, with a focus on topics relating to education and social justice.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Chapter 4 Lone Young Parenthood by Choice? Life Stories in Great Britain**

**Fabienne Portier-Le Cocq**

In 2016 in the UK, there were three million (2.9) one-parent families. Women accounted for 86% of lone parents with dependent children and the remaining 14% were headed by lone fathers (ONS 2016). The rates have only marginally changed since 2001. Two per cent are young parent families.

The country has 19 million families (18.9 m) with dependent children (ONS 2016) and it is estimated that one child in five (22%) is raised in a lone-parent family.

In an initial qualitative study carried out between 2001 and 2005 in England and Scotland through semi-structured in-depth interviews, a hundred young lone mothers were consulted on the circumstances of their pregnancy and their daily lives; what it is like emotionally and practically to be a young parent managing without assistance from the father and/or the family.

It appeared that for some, the relationship with the child's father had ended even before the young woman knew she was pregnant, or the break-up occurred in the first weeks or months of her pregnancy; some others had already made up their minds not to pursue the relationship with the partner or even that the relationship was no longer an option due to the biological father's substance abuse, previous or current convictions, or even abusive behaviour, putting the mother and child at potential risk.

In an ongoing follow-up survey of which the first phase was held from early September 2013 to November 2014 in England, more young mothers or pregnant young mothers being interviewed appear to be living under the same roof with the father of their child, or with another man who has become a 'stand-in father' to the child. Some have an on-and-off relationship, either with the father or with someone new, whilst some prefer, or have no other option, than to bring up the child on their own.

F. Portier-Le Cocq (\*)

Université François Rabelais, Tours, France e-mail: fabienne.portier-lecocq@univ-tours.fr

<sup>©</sup> The Author(s) 2018 75

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_4

The paper will give the latest official figures on young parenthood in Great Britain, and data on the samples of both qualitative studies carried out by the researcher. It will then look at how young lone parents fare financially, and cope emotionally and socially. Other than attending a young parents' group session in a children's centre once a week, where they have the opportunity to meet peers and share their experiences, worries, hopes and fears and can be given advice, help and support, many have no other support network available to them. The chapter will address the concerns they voice.

#### **Introduction**

In 2016, there were 18.9 million families in the UK, of which 12.7 million were married or couples in a civil partnership with or without children, and there were 2.9 million lone parents with dependent children1 . Over the last two decades (1996– 2016), lone parent families grew by 18.6% (ONS 2016). Nearly 55% of lone parents with dependent children have one child, while 13% have three or more. Fathers accounted for 22% of lone parents with non-dependent children (adult children) compared with 10% of lone parents with dependent children (ONS 2016, 5). Less than 2% of lone parents are young parents (Gingerbread n.d.). The official definition of a lone parent in Britain is unambiguous: "A lone-parent family consists of one parent, irrespective of sex, living with her or his never-married dependent children, provided these children have no children of their own" (ONS 2011). In this chapter, whenever the term "young mothers" appears, it means that they are lone parents, namely "young lone mothers". The outcome mostly described in research is negative, despite the majority of lone young mothers of the sample faring well and showing resilience at the very least.

Globally, about 16 million girls aged 15 to 19 years old and some one million underage girls (under 15) give birth every year, which represents 11% of all births worldwide. The majority of these mothers (95%) live in low- and middle-income countries. The 2014 World Health Statistics show an average global birthrate of 15 to 19 year-olds of 49‰, the highest rates occuring in sub-Saharan Africa (WHO 2014). In Western Europe, the United Kingdom has the highest rate of young conception, although, since the implementation of the ten-year action plan (2000–2010) by New Labour government (1997–2010), the rates have fallen and are at their lowest since records began in 1969. In 2012, the conception rate of 15 to 17 year-old women was 27.9‰ conceptions (a decrease of 40.8% since 1969). Most of these conceptions are unplanned, unwanted or the result of sexual violence (ONS 2013a).

In Great Britain, girls who give birth in their teenage years are viewed as a particularly vulnerable group which poses both physical and emotional risks for the

<sup>1</sup>A dependent child is aged under 16 and lives with at least one parent, or 16 to 18 years old in fulltime education. They are not married nor living as a couple and are not parents (ONS, *Families and Households in the UK:*2016a, p. 2, 2016).

young mothers and risks to their children in terms of low school attainment, antisocial behaviour, substance abuse and early sexual activity (The Poverty Site 2014). It has created a stigma of sorts: the poor and working poor lone young parents are viewed a social pariahs, having children in order to receive benefits or see an increase in their benefits.

This paper is based on two qualitative studies carried out in twenty-first century England and Scotland through semi-structured interviews and will endeavour to shed light on the reasons behind young conceptions and births given by interviewees while focusing on the concept of choice to become a parent. Was it their personal choice, which drove them to become a parent? Which factors intervened in their decision-making? It will also present how lone young mothers fare financially, emotionally and socially; the support, if any, they are being given within their own family or elsewhere. Lastly, the article recounts the issues young parents are confronted with and the concerns they voiced.

#### **Methodology**

Britain has been notorious for having the highest rate of young conceptions in Western Europe as well as more children living in poverty compared to other European member states, and young motherhood is considered to be a social or a public health problem. Giving a voice to young mothers and pregnant girls through qualitative studies has been the primary aim of the researcher, as the insiders' views are almost always absent from scientific literature. The corpus consists of two qualitative surveys, both conducted solely by the author. During the first one, between 2001 and 2005 at thirteen locations in England and Scotland, over one hundred young mothers or young mothers to be were interviewed. The questionnaire was composed of fifty questions under various entries. Agreement was made prior interviews that informants were free to not answer questions, should they deem some too personal and intimate given the sensitivity of the subject matter. The gathering and recording of conversations took place on a voluntary basis with respect to ethics and confidentiality (Portier-Le Cocq 2009). A large majority of lone women were interviewed on their own; very few asked for a third party to be present: a social worker, a midwife or a female friend to attend. No incentive was provided in return to the participants or their children. All interviews were recorded entirely and transcribed verbatim and the content analysis of seventy-eight interviews was performed. The underlying reasons for the attrition rate are occasional defective recording equipment or noisy environments, such as babies crying, children playing loudly, or some young mothers not being forthcoming on the issue. Two-thirds of the interviews took place in England.2 Respondents were between 12 and 19 years of age at the time of conception and 60% were underage. To the best knowledge of the author,

<sup>2</sup>The author was unable to undertake a qualitative study in Wales as she was told repeatedly that young parents were too vulnerable there.

the age of consent in the United Kingdom being 16 (Sexual Offences Act 2003 and the Sexual Offences (Northern Ireland) Order 2008), the data gathered are unprecedented and innovative, if not unique. It is unlawful for an underage girl to engage in sexual intercourse – any type of sexual activity with a person under the age of 13 is strictly illegal – and a child protection issue. The first phase of the second ongoing study was conducted from September 2013 to December 2014 and commenced in the south-east region of England with thirty young parents in children's centres. Some facilities organise specific weekly sessions for parents under 25 years of age.

The scale of the sample is rather impressive if one accounts for the fact that most findings in connection with young motherhood in the United Kingdom are customarily based on fifteen interviews. Research drawing on the content analysis of a sample of one hundred and eight semi-structured interviews shows that this segment of the population is not a homogeneous group. Gathering first-hand information by meeting young mothers, giving them a voice and exploring the circumstances of these women who were marginalised by British society was the starting point of the research. A second aspect aimed to dispel the myths surrounding juvenile pregnancy and motherhood by comparing the young mothers' narratives with scientific literature and political rhetoric. In effect, the rationale put forward by successive Conservative governments in relation to teenage pregnancy and motherhood was that girls were pregnant on purpose either to access council housing or to claim and live on benefits. The reality emerging from both these large-scale qualitative studies is far more complex. Bringing out the causes and consequences of the pregnancy and understanding the reasons to carry on the pregnancy at this phase of life were further research objectives. Studying the repercussions for mothers, children and society at large was another purpose, together with looking at their living conditions on a daily basis.

The rate of young parents can seem paradoxical in a country which pioneered contraception3 , where abortion was legalised in 1967 (Abortion Act 1967)4 in Great Britain, and the Family Law Reform Act 1969 enacted that a minor over 16 could get a termination without parental consent.

#### **Lone Young Parenthood and Its Political Approach: A Success Story**

Young childbearing is a very complex issue whose causes and risk factors are multifaceted and whose outcomes can sometimes prove positive. In Great Britain, it is seen as a problem, notably when the mother is not married and no support is provided; and it is also associated with social and physical consequences which have public and private costs and have negative effects on the mother and her child.

<sup>3</sup>The birth control pill has been available since 4 December 1961.

http://news.bbc.co.uk/onthisday/hi/dates/stories/december/4/newsid\_3228000/3228207.stm

<sup>4</sup>The Abortion Act 1967 came into operation on 27 April 1968.

According to some, it is thus a social, economic and public health problem in the case of unmarried mothers. Defining this phenomenon as a problem marginalises and stigmatises young lone mothers and their children all the more so as quantitative surveys are more frequent than qualitative ones and they do not differentiate between wanted from unwanted pregnancies. As Maureen Freely wrote in the introduction of one of her books *"It's the statistics on divorce, illegitimacy, teenage pregnancy, delinquency, crime and urban blight that prove the family is in trouble, can no longer perform its social functions, needs to be saved, improved, supported, warned or admonished"* (Freely 2000).

Being an underage mother ranks first in social exclusion, together with expulsion from school and homelessness. These different forms of exclusion affect 1% of the population (SEU 2001). The social exclusion notion reflects the hindrances people who live in poverty or in deprived areas are confronted with, and which prevent them from participating in society. When Tony Blair was Labour Prime Minister (1997–2007), his government implemented the Teenage Pregnancy Strategy, and one of the priorities was sexual health (The Stationery Office 1998). He commissioned a report by the Social Exclusion Unit into the causes of these problems and possible strategies. The report (SEU 1999) released in June 1999 disclosed that each year 90,000 English teenage girls (Graham 2004, 127)5 were pregnant, about 8.5% of whom were underage (7700) and 2.5% were under fourteen (2200). Sixty per cent of conceptions resulted in live births. The government pointed out that twothirds of girls under sixteen had not had any sexual intercourse and that most had never been pregnant before turning 20.

The two key stages of the Labour government Teenage Pregnancy Strategy (1999–2010) aimed to halve the number of under eighteen year-old conceptions in England. An intermediate objective by 2004 sought a decrease of 10 to 15%, depending on the geographical areas, but was only partially reached despite joint efforts (11%). Variations reflect the characteristics of deprivation in England. Half of all early conceptions take place in the 20% of the most disadvantaged wards, usually inner cities, and young mothers are more likely to come from a lower socioeconomic background and a large family, with parents showing little interest in their education, and/or living in a lone-parent family. Data show that girls from loneparent families whose mother did not work outside the home were more likely to become young mothers than those whose mother worked outside the home (Botting et al. 1998; Portier-Le Cocq 2009). The second goal was to increase the participation of young parents aged 15 to 19 in education, employment or training by 60%. Fifteen per cent of all young people who are not in education, employment or training (NEETs) are young mothers or young pregnant girls (Local Government Association 2013); the proportion of 16–18 year-olds who were not in education, employment or training in 2011 stood at 10% (Hagell et al. 2013).

<sup>5</sup> It means that '*in England and Wales, only about one in thirty-five young women aged 15–19 year have a baby each year, and about one in fifty have an abortion. The rates are much lower in younger teenagers. Only about one in 250 13- to 15- year-olds have an abortion and about the same number have a baby each year'*.

In Scotland, the national sexual and reproductive health strategy has aimed to decrease the rates of unwanted pregnancies. Reducing by 20% the conception rate of 13–15-year-old girls was the main concern, especially in disadvantaged areas. The fertility rate of 15 to 19-year-old girls was 29.3‰ in 2000 compared to 17.9‰ in 2013 (General Register Office for Scotland 2015).

According to the 1999 report on teenage pregnancy, three key-factors stood out explaining the causes and can be summed up in one word: neglect. The first is low expectations (young people see no prospect of a job and fear that they will end up on benefit); the second is ignorance (young people lack accurate knowledge about contraception, sexually transmitted infections, what to expect in relationships and what it means to be a parent); the third is mixed messages - it seems sex is compulsory but contraception is illegal. The adult world bombards teenagers with sexually explicit messages and an implicit message that sexual activity is the norm and parents and most institutions are silent or embarrassed when it comes to talking about sex. 'If sex is not talked about, it won't happen' (SEU 1999, 7).

There is also a strong economic argument, as the costs of income support benefits for early pregnancy and motherhood was £26 million in 2009–2010 (TPA 2011) and the NHS costs associated with births, miscarriages and abortions were estimated at £63 m a year (TPU 2003). Young lone mothers are perceived as a burden on the taxpayer, especially because they have fewer skills and opportunities for finding a job. They either choose to, or have to, drop out of school early and more than half never resume their education, even though they are below the statutory school leaving age, and therefore they have fewer or no qualifications and are of a lower level than their peers. Factors attributed to a high early pregnancy rate are low aspirations, lack of self-esteem, behavioural problems, lack of appropriate services, and a culture that supports young motherhood. The findings of the author's research show there is a myriad of reasons: poor family relationships, lack of communication, lack of early information on sexuality and contraception from parents, inefficient and inconsistent sex and relationship education in schools or the subject being broached too late, binge-drinking, a lack of self-respect and dislike of school all contribute to early pregnancies.

It is indeed a success story since funding had been ploughed into the ten-year action plan and that the policies implemented by New Labour with respect to the reduction of teenage conceptions have proved successful and fruitful. The conception rates for under 18-year-old women have been the lowest since 1969 – the first year for which abortion data was officially recorded – and the rates have kept falling. The same applies for under-16-year-old-conception rates, with a fall of 10% between 2013 and 2014, for instance (ONS 2016).

#### **Findings: Some New Insights**

From the findings of the author's field research, there appears to be a discrepancy between real practices and stated practices. This study differentiates itself from other previous studies for three main reasons: the age of the mothers interviewed, the scale of the study, and the themes broached – the sensitive issue of the sexuality of lone young mothers has been largely overlooked in other surveys.

Young motherhood is seen as unacceptable: young lone mothers are undesirable and successive governments have done their utmost to present them in this light, although not all early pregnancies are problematic (Duncan et al. 2010). Public opinion, the media, politicians, and some academics, devote themselves to marginalising, stigmatising and constructing early motherhood as deviance, in the same way as alcohol or drug use and abuse, violence or delinquency (Portier-Le Cocq 2014). Nonetheless, for some young people, a child is an acceptable choice of life because it makes sense and values young parents (Cater and Coleman 2006, Portier-Le Cocq 2009, 2014). Except for the age at which they fall pregnant young mothers see no differences with older mothers (Sellenet and Portier-Le 2013). Besides, the medical profession has revised their opinions and shown that being young and a mother is not an obstacle on the health and obstetrics fronts (Portier-Le Cocq forthcoming). Age is important, as the possibilities and rights differ if the mother is under or over sixteen. Being an underage mother means that they are not allowed to claim nor to receive any state benefits. New Labour was the first to tackle the issue of early pregnancy within the frame of social exclusion. Its main concern was to include young parents. In England, inclusion since 2003 has meant that young pregnant girls or young mothers – from the sixth month of their pregnancy who cannot live at their parents' or their partner's home have to live in supervised semi-independent housing.

Among the explanatory causes of juvenile pregnancy given by the author's sample, binge-drinking or the excessive consumption of alcohol was occasionally mentioned. In this case, it was unprotected sex; some did not recollect having been sexually active and the denial of their pregnancy could vary. The specialised literature and the policy-makers prefer to put forward promiscuity. Young mothers are often depicted as irresponsible and immature girls. However, these young women most often fall pregnant the first time they have sexual intercourse because they are pressured by friends into having sex to belong to the group, because of casual or non existent contraception use (SEU 2001, 49)6 or are under the influence of alcohol. They have a very low awareness and poor access to emergency contraception (Braid 2008), which respondents hardly ever mentioned. There is also an assumed feeling of invulnerability - "it won't happen to me" (Lavielle 2000). The lowest levels of contraceptive use are found in the 16–19 year-olds (Hagell et al. 2013, 63). Twothirds (64%) of 16–19 year olds are "at risk" of pregnancy – they have a sexual

<sup>6</sup> '*Between a third and a half of sexually active teenagers do not use contraception at first intercourse*'. The proportion is higher than in many European countries.

partner and do not use any protection - but one in ten of those with a partner do not use contraception (Lader 2008–9). Peer pressure, curiosity, conformity with others, because they feel excluded from the group if they are not sexually active are critical vehicles of pregnancy at an early age.

If we consider the negative consequences brought to light by the young lone mothers in both qualitative studies, many wished they had deferred sexual intercourse or, when they found out that they were pregnant, they had kept to their initial decision to have an abortion. Many reasons, when they had the choice, were put forward to decide against a termination: the fear of the surgical procedure, the details provided by the health professionals pertaining to abortion, the baby's biological father's or his family's pressure on the young pregnant girl, pressure from the young girl's parents, social class, and the type of antenatal examination. As far as the baby's father is concerned, a vast number of young mothers claimed that they deliberately ousted him during or after pregnancy, when it was not the young girl's parents' decision. In the corpus, unlike Scottish girls, many English young pregnant girls were thrown out of the parental home as soon as they broke the news to their family. Also, some young men shirk their responsibilities at the disclosure of the pregnancy or terminate the relationship without being informed of the pregnancy.

Contrary concerns expressed include tiredness, lack of time and energy, isolation, insufficient state benefits to live on and to bring up their baby decently, an unbalanced diet, no longer having a social life, not going out with friends or elsewhere because of the lack of money or the lack of transport or its cost for that matter. On a positive note, the author has observed a change in the decade between the two qualitative studies: a notable decline in the stigmatisation of young lone mothers. Compared with the interviews conducted in 2001–2005 when young mothers described judgmental attitudes from the general public, practitioners and services, in the most recent study in 2013–2014, no mothers reported negative and derogatory comments.

Against all odds, young motherhood can sometimes be positive and unproblematic (Scottish Parliament 2013). Relationships with the family, especially with the mother's own mother, improve with the birth of the child. The family then supports the mother and her baby financially and emotionally and they also give her advice. The young person, who was most often disaffected from school, expelled or had dropped out, resumes her studies or starts some training. The main reasons given were that she wanted to manage on her own to bring up and educate her child and no longer depend on state benefits, which are deemed insufficient. Above all, she wants to prove to others that she is a good mother and a responsible young woman, and that she can make do; she wants her child to be proud of her in the future. Mothers make plans for the future, which are often imposed by financial issues. They want to find a job so as not to depend on benefits; to look after their child and spend a lot of time together, seeing them grow up. This is inconsistent with the government views, as lone young mothers feel guilty when they work or attend school since, in their opinion, they cannot be with and take care of their child enough.

#### *Young Mothers Speak Out!*

Drawing on the qualitative surveys conducted by the author, four main causes of early motherhood emerged, which are personal, social, economic and environmental (family). The causes and risk factors include risk behaviours, education, family and social circumstances. The family and the community disengagement are essential agents.

When summarising risk factors with respect to the sample, they encompass poverty, being in or leaving care whether it is in an institution, a foster family or with adoptive parents, casual sexual relationships, being the daughter of a young mother, living in a large family, and not being in education, employment or training (NEET). Underachieving at school before the age of sixteen without any qualifications and living with parents who show little interest in their education are other causes together with mental impairment, psychological problems, low self-esteem, peer pressure and sexual abuse or rape in their pre-teens.

"How old were you when you first had intercourse? — Um, about (thinking) 10. I was forced into it.

Was he older? — Yes. He was about 35.

You were raped… — Yeah, but I have told none, except my boyfriend, so… He's in prison now.

How old were you when you first consented to have sex? — Um, if you do not include the rape at 10, I was 12 or 13."

*Mel,*<sup>7</sup> *nearly 16, mother to a six-month-old child.*

"When did you become pregnant? — Um, actually I was raped in my country, that's why I came here as asylum seeker so I was pregnant in October, but I didn't know that, cos I was raped in August. So I didn't know until January the following year that I was pregnant.

How advanced was your pregnancy? — Um, I was, I think six months pregnant. We went to another organisation and they said it was too late for me to have an abortion, so they said it's risky.

Would you have had an abortion otherwise? — Um, in my country they don't recommend that, because in my country if you do that they put you in prison for at least a year or something, yeah. But in my situation, maybe I could do that because I didn't have any choice because of my age. I wanted to study; I could do that but it was too late for me to do that. And also I just found that I was having twins so I could have done that.

Did you consider having the baby adopted? — No, I couldn't do that. They told me about that but I couldn't give my baby to someone else."

*Anna, 17, mother to a thirteen-month-old baby.*

Early motherhood is the result of a complex conjunction of factors, one of which is growing up in poverty. However, the most prosperous areas also have higher early pregnancy rates than the average of the European countries. The country's geographical areas along with demographics are important and there are geographical

<sup>7</sup>All names have been changed.

variations too (Arai 2009, 25). In 2000, the juvenile pregnancy rate in Richmondupon-Thames was 19,4‰ while it was five times higher (89,8‰) in Hackney. Young parenthood concentrates in areas characterised by social exclusion and deprivation (mining, former industrial and seaside areas).

Coastal and some hinterland areas show conception rates, which are higher than the national average. The reasons are isolation in rural areas and the fear of being seen and talked about when/if they buy contraceptives at the pharmacy counter, whereas on the coast, leisure activities and business give a sense of hedonism (Bell et al. 2004). As an explanation for this statement, the carnival atmosphere due to fairs, the pier and video arcades are propitious opportunities and venues to meet seasonal workers and people in general, hence transient relationships. To cap it all, casual unprotected sex, alcohol and drug consumption, and abuse are further elements. Furthermore, researchers reached the conclusion that by working in the leisure industry under sixteen year-old girls were accustomed to attracting older men in pubs and clubs and that youths were under the impression of being used by the demands of the local economy.

Young mothers have lived in broken families, reconstituted families, poor household or lone-parent families predominantly headed by mothers. They very often moved out, hence changes of schools and a poor level of education for mothers or their own mothers. They had sexual intercourse when very young - sometimes not of their own accord (CSJ 2014; Barter et al. 2009; Coy et al. 2013) - and either did not use or only occasionally used contraception. Three-quarters of the interviewees had had sexual intercourse before the age of sixteen, and half of the sample said their pregnancy was unplanned. Half of the sample said they did not use contraception when they first had sex.

"How many boyfriends have you had so far? — I started having sex at 11 with older boyfriends, like 15."

*Chantelle, 22, mother to a nineteen-month-old child and a four-and-a-half-year-old child.*

Alcohol abuse or binge-drinking, which has been ignored by government surveys, was a factor mentioned a few occasions, and most often the young women had no recollection of what had happened as both extracts below demonstrate.

"I can't remember. I was drunk when I had sex, Jesus Christ! I fell pregnant when I was thirteen and that was the first time I can't even remember." *Daisy, 16 and two months, mother to a two-and-a-half-year-old child.*

"How old were you when you had your first sexual intercourse? — 14. Um (laugh), um, I cannot remember, I was kind of drunk." *Jasmine, 15 and two months, mother to an eight-month-old child.*

The Institute for Public Policy Research also stated that early pregnancy was part and parcel of British culture. Researchers clearly showed that British young people were sexually active at a younger age than their European counterparts and that they had more casual sex. This can be explained by the increase in time spent with their peers in non-structured activities such as hanging out in the streets and not being under their parents' supervision, especially in deprived communities rather than time spent with parents (quality time). Idleness and a feeling of a lack of parental control are sources of early pregnancy.

The young British are supposedly more interested in consuming and spending long hours on the Internet without any parental control, and research shows that they are exposed to more sexual messages via the use of the Internet or in the media (Margo et al. 2006, 171–72). These messages promote sexual activity but never allude to contraception or HIV transmission risks. According to Carrys Lovering (2005, 134–35), young people wish they were busier and had venues at which to meet (Barry 2005). Some young women wished that youth centres had longer opening hours, that they were informed of the planned activities, and that trips and artistic projects and information technology training were organised as well. Young people are keen on outdoor activities such as skateboarding, for instance, but they said they had nowhere to practice and grown-ups complained if they skated in the streets.

Monica Barry said that youths from deprived communities and, especially the young unemployed, do not go to leisure centres geared to them because of the cost. As a consequence, groups who need this are those who are deprived of this amenity for economic reasons. She also said that young mothers go to young mothers' groups such as *Young Women Christian Association* because they are free. If there were no such services or Sure Start children's centres, for instance, lone young mothers would not meet anyone, above all because of the lack of childcare or its cost.

Young women have nothing to lose if they get pregnant, as they are often disaffected from education and their achievement is low as well as their career expectations. Becoming a mother is a way of being acknowledged within one's peers, family circle and society, and of reaching adulthood. Having no qualifications, being disengaged from school and having no career prospects, having a baby is thus an acceptable option. Social background and class are crucial factors in the decisionmaking process. Young girls who come from affluent backgrounds and aim to carry on with their studies at university are seven times more inclined to have a termination than their less affluent counterparts (BPAS 2000).

#### *Living on a Tight Budget: The Big Issue*

State benefits provide only enough for survival and young lone mothers struggle to make ends meet. Receiving income support, which is a means-tested benefit, is an indicator of poverty in the UK. Compared to older mothers, who had a child at 24 or over, the percentage of young lone mothers who receive income support is three times more. The Department for Work and Pensions figures show that around 70% of mothers aged 16–19 claim income support. Most of the time the young lone mothers of this study had no idea which benefits they could claim or what their parents received for them and their children, which shows they did not get pregnant to get money from the State (Allen and Bourke-Dowling 1998).

"Do you get any benefits from the state? — My mum claims.

Do you know what she gets? — Um, she gets a £500 grant for us to get all the baby stuff, and I think she got money for clothes for us as well; when you get bigger you need clothes. I don't know what'll happen when the baby's born, and what the money I'll get, I don't know."

*Debora, 15 years old and three months, mother to a 4 month-old child.*

Young lone mothers may be entitled to benefits depending on their age, whether they are living with their parents or not, and if their parents work or not. Sixty per cent of underage mothers were not entitled to claim any benefits. They were thus dependent on their partner, if he was still around, or their parents if they had not forced them to leave home. They were only entitled to weekly Healthy Start vouchers to help with the cost of baby milk, milk, fruit and vegetables. You must be 16 or over to claim Income Support, which is allocated to lone parents who are not working (or working fewer than 16 hours a week) and whose youngest child is under five. The map of young pregnancy in Britain is the map of poverty and deprivation.

Another crucial issue for young lone mothers is that childcare can prove nonexistent or onerous and hinder the process of getting young mothers into work.

"I could [stay at school] as long as I had childcare but that was the problem because we didn't have a lot of money to pay for a child-minder to look after the baby when I went to school, but then it would be a problem with coming home with homework and having to look after the baby at the same time because you don't get free work periods or free time or teachers at that school. There was nothing really there for me that they could help me with."

*Helen, 15 years 10 months, mother to a 9 month-old child.*

Hilary Graham and Elizabeth McDermott (2006, 30) underlined that families help ensure that their daughter and her child are provided for materially and with a home, among others. In the case of the breakdown of the links following the announcement of the pregnancy, it is very difficult for young mothers to overcome the financial and social handicaps of being a young and poor lone mother. If the young couple has separated and they have built up a friendly relationship, the baby's father can provide the young mother with space, support and things for the baby instead of money.

"He takes Mike like all day on Sunday but he doesn't help financially or anything.


So, how do you manage financially? — He helps us all a lot.

He gives you maintenance? — Yeah. No, he just gives it straight to me. He helps out with buying nappies and things like that and then he gives us money for when I go out."

*Fiona, 17 years 10 months, mother to a two year-old whose 21-year-old father is a factory worker.*

"Almost every day like he comes from work to my house. He stays for a bit, but if not everyday he takes the baby to the crèche on Thursdays and Fridays in the morning. I have to wake up early, and he comes on Sundays to take it. He goes to his mum's so he takes it with him, and sometimes during the week, he just picks the baby up from the crèche, so I wouldn't be able to say exactly how many times.

Does he help you financially? — I'm always asking for things, then he helps.

Every week he gives me money for the baby's crèche and other things.

Still now that you split up… — Yeah."

*Megan, 19, mother to a 9 month-old baby whose father is 24 and a receptionist.*

#### *A Roof Over My Head*

It is generally not true that girls get pregnant in order to get a council flat; 90% of young mothers live in their parents' home or the home of their partner's parents and there is also evidence of planned juvenile pregnancies.

The thrust of the New Labour Government's message was that young parents needed support to maintain a tenancy and live independently. This includes teaching them how to cook and how to budget. All lone young parents under 18 who cannot live with their parents or partner should be in supervised supported accommodation and they should not be given any independent tenancies without support. In July 2013, the 40 Group - the Tories who hold the most marginal seats in Parliament contemplated depriving young lone mothers of council housing or housing benefit in a bid to reduce early pregnancy. All benefits would be conditional upon young lone mothers living at the parental home or in supervised hostel accommodation.

In the sample of the first qualitative study, 50% of mothers lived at their parents' or relatives'. One-fifth lived in independent housing, two-thirds of whom lived in supported accommodation with or without the child's father. The discrepancy between England and Scotland in this respect is that early pregnancy and motherhood seem to be more acceptable in Scotland, especially in the most deprived area, or the Scots seem more understanding or more responsible parents. Moreover, in the sample, no young pregnant girls were asked never to return home as was often the case in England. Young Scottish mothers tend to live at their parents' until the baby is a toddler then move to their own independent housing with their child. The lodging always has to be close to their parents' home and they go and see their parents on a daily basis, which is reminiscent of M. Young and P. Willmot study's findings (1957).

"Just up the road from my mum and dad's." *Sara, 17 and 9 months, mother to a 9 month-old baby.*

"Yes, across the road from my mum.

[…] when you had the baby did you live at your mum's or… I stayed with my mum for a year but then I moved out 'cos like, my sister, Muriel, my sister lived there as well and she's had a baby as well, so it was really overcrowded. So then I moved out, got my own house."

*Kirsty, 19, mother to a three year and 6 month-old child.*

Overcrowded housing is also a reason for having to live on one's own in council housing.

"Just, I have to scrape it together… I live on me own with Davy. I have moved from my mum and dad's house. I'm getting House Benefit.

And how much is it? — Um, £42 a week.

And is it OK for the rent? Sometimes.

Do you get Child Benefit and Income Support as well? — Yeah. £30 on nappies and wipes and baby food, £30 on gas and heating; so it doesn't leave a lot of money for food for me or clothes or so…

And why is it you left your parents' home? — Because there was eight of us and there are three bedrooms. So there wasn't enough room."

*Sara was 16-and-a-half at the conception of her baby who was 9 months old at the time of the interview.*

Young lone mothers very often ask to live in the vicinity of childcare facilities so as to avoid spending money, time and energy on transport. David Pevalin (2003, 8) said that young mothers' children were the first to be affected by residential mobility, i.e. they could live in four different places or more before they turn ten.

#### *Did You Say Social Life?*

Preferably, young lone mothers should not move out. They have to organise their new life as mothers, they live on a tight budget and have at least one child to nurture. They are sometimes put up in temporary accommodation such as bed and breakfast, a mother and baby home or accommodation for the homeless. Once they have a roof over their head, their mobility is limited, they live in poverty, their freedom and social life dwindle and they spend long hours on their own trying to entertain themselves and their baby in complete isolation. They are confined to their home and their close neighbourhood, which could explain their wish to live near their family, which in turn increases the difficulties in finding accommodation.

As for their social life, despite the problems, half of the sample said they went out regularly, but 15% had never been out since the baby was born, which could have been for months. Ten per cent seldom went out because of the lack of money or tiredness. Social activities may include going to the pub, drinking or binge-drinking, eating out, hanging out with friends, going to the pictures and going food-shopping. For those who have no social life at all, watching television is their only outlet.

"— I don't go out quite a lot. I'm quite happy watching some TV." *Abigail, 19, mother to two-year-old twins.*

"There's a park just where I live where we meet all and we might go to the pictures or go for some tea but there's just times we are just all together so that I'm making most of the time while I've got it."

*Helen, 15 years and 10 months old, mother to a 9 month-old baby*

Over time, visits and going out with friends or relatives become a rarity as they lack money, time, energy and support, and young lone mothers have a different pace of life from their friends when they become parents. Isolation is what sometimes makes them regret having a child. Still, all the negative points previously developed are offset by their pride and fulfilment of being parents. Childbearing and childrearing are turning points in young women's life for the better. With maternity, they have a goal in life, more self-esteem and are now respected by their parents who consider them adults. Young lone mothers can and want to rebuild their lives through their resilience and in order to show others that they are not the inadequate or failing mothers they are alleged to be. Yet they do need emotional and social wherewithal to achieve that.

#### **Conclusion**

This chapter has aimed to call into question theory in young mothers' studies through systematic content analysis and young mothers' everyday lives as far as their social and private lives are concerned. Despite positive and life-enhancing attitudes to motherhood, what they lack most is support and back-up, encouragement, money and social network. A striking comparison between the two qualitative studies conducted by the author has so far shown that young fathers in the south-east of England are more involved in "being there for" the mother and the child. They support their family by working, living under the same roof, attending young parents' sessions in Sure Start family centres, and caring for the mother and child.

The interviews run counter to generally held views and the official line of the Conservative Governments that young British women are scheming to have a child to jump the house queue and live on state benefits. In Great Britain and the other industrialised countries, the perception of young lone mothers is negative partly because they drop out of school and then have difficulty in finding a job. The blatant lack of childcare and its prohibitive cost is another barrier to their entering the labour market. For the majority of young women who grow up in a country of the OECD8 , the norm and the expectations are to study, have a job with a household and two salaries, defer a pregnancy then have a small family. In this respect, early motherhood has become a significant issue because there is a discrepancy between the majority of the population and the young mother who is seen as different (Sellenet and Portier-Le Cocq 2013). Moreover, even if the early pregnancy rate has seen a 41% reduction in under-18 conceptions and an even steeper reduction in births by 46% since 1998, the major change since the 1960s has been the high number of babies born outside marriage, leading to more one-parent households in Great Britain as well as the rise of house-sharing since the 1980s. In 1981, 55% of births involving young mothers were within marriage compared to 12% in 1996 (Botting et al. 1998). Sixty per cent of young mothers are lone mothers (DCSF 2007, 8). Girls who become mothers, either no longer attend school, are expelled or disaffected. At the birth of their child, most only want to care for their child and are reluctant to entrust the child to the care of a third party. The determination of past and present governments to put young mothers into work, education or training is incompatible with that perception. Another paradox lies in the fact that governments want to encourage young lone mothers to spend time with their child while also encouraging them to find a job so as to be less of a financial burden to the taxpayer.

<sup>8</sup>Organisation for Economic Cooperation and Development.

The causes and outcomes of juvenile conceptions and motherhood are a cause for concern in Great Britain, but to the outsider the main efforts of governments concentrate on preventing young motherhood, while, it seems, passing over in silence disturbing and upsetting factors such as binge-drinking among youth and the discrepancy between income distribution and inequality at large, which are altogether paramount. Assistance brought to young lone parents in terms of support had thus far not been a priority on the political agenda. Yet it should be borne in mind that the children of young parents and young parents themselves are tomorrow's citizens and taxpayers and that they should not be scapegoats.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **References**


**Fabienne Portier-Le Cocq** is Professor of Contemporary British Studies at the Faculty of English Studies at the University François Rabelais of Tours, France. Her research interests include teenage parenthood in Great Britain and France; lone mothers, motherhood, violence against women and girls, care leavers, social issues, qualitative studies and comparative European studies.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

### **Chapter 5 Variety of Transitions into Lone Parenthood**

**Laura Bernardi and Ornella Larenza**

#### **Variety of Transitions into Lone Parenthood**

The rise in lone-parent households is part of the growing diversification of household types, living arrangements, and family forms that have presented in Europe over the last 40 years. In 2009 in Europe, the share of lone-parent households (calculated as the proportion of households with children under 18) ranged from 5–7% in countries like Greece, Spain, Romania, and Slovakia to 20–24% in countries like Estonia, Latvia, Ireland, and the United Kingdom (Chzhen and Bradshaw 2012, p. 490). In Switzerland, the last available survey data from 2013 indicated that 15% of households with children under 25 had one parent. More than 80% of these situations regard lone mothers (2015).1 Despite lone-parent households representing a minority of households, researchers are interested in them for two reasons. On the one hand, lone-parent households are considered at risk and are often mentioned when talking about family risks, particularly economic and health risks (Bonoli 2005). On the other hand, they are an alternative living arrangement increasingly contributing to the family diversity that characterizes contemporary families.

Research on lone parents in Europe, which rests on analyses of nationally representative, large-scale surveys, has shown that lone parents are different from other parents in terms of socioeconomic and health characteristics. Lone parents have higher risks of negative outcomes (poverty, unemployment, health) than parents in couple. This is particularly true for lone mothers but less so for lone fathers. Lone

<sup>1</sup>These figures refer to the prevalence of lone parents in the population at any given point in time. Yet, if one were to have data from a longitudinal perspective and could calculate the percentage of individuals who have ever been single parents, it certainly would be higher, meaning that this condition is experienced and relevant for a larger share of the population.

L. Bernardi (\*) • O. Larenza

National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life course perspectives (NCCR Lives), University of Lausanne, Lausanne, Switzerland e-mail: Laura.Bernardi@unil.ch

parents are also much more likely to be social assistance recipients or to benefit from social housing and health insurance benefits (Whitehead et al. 2000). Children in one-parent families have a much higher risk of living in poverty or social exclusion than dependent children in two-adult families. Around half of one-parent households with dependent children were at risk compared to only about one-fifth of households with two adults and two dependent children (Lopez Vilaplana 2013). Lack of resources (financial but also psychological and social) and a limited capacity to recover from stresses in other life domains (work in particular) are factors of lone parents' vulnerability. A number of qualitative studies devoted to lone parents, especially in France, the UK, and the United States, have shown that besides the vulnerability associated with lone parenthood, there exist important aspects of resilience triggered by the lone-parenthood experience. In the Anglo–Saxon context, teenage lone mothers from disadvantaged backgrounds hold a positive attitude toward motherhood as turning point in life that allowed them to take a different direction, including going back to education and work (Bell et al. 2004; Coleman and Cater 2006; Duncan 2007; Edin and Kefalas 2011; Phoenix 1991; SmithBattle 2000). Lone mothers and parents who coupled represent equally positive parental role models (Dowd 1997). Research in continental Europe has additionally brought to attention that the population of lone mothers is much more varied than its perception as simply a population composed only of low-educated adolescent mothers. Such research has consistently shown that lone mothers are not only overrepresented in social welfare compared to mothers in a couple but also are more likely to work, and to work full time, than mothers in a couple (Struffolino et al. 2016); it may be of interest to researchers that lone mothers are more likely to reside in poorer districts but also in more well-off ones than mothers in a couple (Lefaucheur and Martin 1997).

What emerges clearly from previous research is that lone parents constitute a very diverse and heterogeneous group and much of this diversity depends on the way lone parenthood is produced and when it occurs in the life course. The most common pathways to becoming lone parent are through divorce or separation, widowhood, and pregnancy or adoption by individuals not in a couple. While widowhood was the privileged path to lone parenthood in the past (Kiernan et al. 2004), with growing union instability, marital break-ups are the primary cause of parents raising children alone for some time in their lives. Besides material and health deprivation, family researchers are interested in lone parents because they represent a nonnormative way of being parents that affects a growing number of children, at least during part of their childhood. Lone parenthood is nonnormative not only because it concerns a minority of parents (although it is a growing minority) but also because shared norms about parenthood in Europe still largely indicate that two parents and children living together is preferable and more appropriate. This is also because being a lone parent is only rarely a planned and a chosen way to parenthood; more often, it is an unexpected or unintended turning point in life.

The lone-parenthood experience itself varies depending not only on the material and social resources available to parents and children but also on the specific legal arrangements with the nonresident parent (provided that he or she is living and available), such as arrangements for child custody, parental authority, and child

support. The age and the number of children are important aspects related to this experience. Although quantitative data and analyses are certainly necessary to gain an understanding of family structures and processes that can be generalized across populations, qualitative studies are critically important to formulating new hypotheses and theories about new family forms. In-depth knowledge from a few cases can be used to do so (Bernardi and Hutter 2007; Ragin and Becker 1992). Analyses of qualitative data provide us with answers regarding how some processes take place, which is critical to theory development. Last, the measurement of lone parenthood in surveys requires standardization; we know that estimates of the prevalence of lone parenthood are affected by how it is measured (Letablier 2011).

For all these reasons, we aim to contribute to the understanding of lone-parent families' experiences through a qualitative study based on semistructured interviews. In Switzerland in 2013–2014, we interviewed women and men who had recently faced an experience of lone parenthood with full custody of their young children. We focused on the transition to lone parenthood and its boundaries from a life-course perspective. In the life-course tradition, *transitions* are defined as periods of rapid change in the life course when individuals redefine one or more of their social roles and pass from one phase to another. Transitions may represent turning points in life when they challenge and redefine future expectations and trajectories in fundamental ways. The process is not necessarily unidimensional and unidirectional, and a qualitative study is crucial to pointing out the challenges of identifying ways to model and measure lone-parenthood trajectories appropriately and improve the quality of our data about this increasingly important family form. At what point can we consider lone parents as such? Is it at separation due to discord, at formal residential separation, or at the moment when the financial and legal responsibilities for the child are assigned to one parent? Through this examination of the transition into lone parenthood, we aim to contribute to the understanding of what events and circumstances can be best considered as defining the onset of lone parenthood.2

#### **Background and Significance**

Who are lone parents? Lone-parent households are not easily and univocally identifiable since their definition varies from country to country and from data source to data source. The most comprehensive definition includes households where one parent coresides with his or (more often) her children and bears the financial responsibility for the children alone, irrespective of whether other adults coreside in the household. Surveys often have more restrictive distinctions, which exclude situations where a parent resides with his or her children and a new partner, or with his or her children and their grandparents, other family members, or unrelated adults. Other criteria, which may vary, are a child's age (some data sources limit it to

<sup>2</sup>Similar questions can be asked about the definition of the end of the lone-parenthood state and constitute our next investigation.

age 16, 18, or 25 to qualify a household as a lone-parent household). In Switzerland, the official statistics of the Federal Statistical Office recently distinguished the households in which at least one child under 25 lives with one parent alone as a specific living arrangement (Mosimann and Camenisch 2015). Other than this official descriptive criterion, there is no legislation addressing lone-parent households as such.3

Research dealing with lone parenthood in demography is concentrated on lone parents as defined by the type of household or living arrangement. More recently, and thanks to the availability of longitudinal data, pathways to and out of lone parenthood have concentrated on the typology of individual union and employment trajectories before and after lone parenthood (Bastin 2013; Schnor 2013; Struffolino and Bernardi 2017). Most of these studies identify a date of entry into lone parenthood as if there were a straightforward time marker for it, ignoring the complexity and the number of events involved in the transition process.

Yet, the onset of lone parenthood is likely to be a fuzzy process. The difficulties in identifying lone parents emerge from at least two distinct sources: one relating to the partnership status and the second relating to the parenting responsibilities and practices. First, the kind of noncoresidential partnership may vary depending on the arrangements between the members of a couple or the financial and emotional involvement of the partner with the children. Consequently, establishing the extent to which lone parents are "alone" is an empirical question. Second, independent of the possible involvement of a new partner, a second set of difficulties in the identification of what lone parents are comes from the assumption of parenting responsibilities and the residential and care arrangements for children. When children circulate across two parents' households according to more or less fixed schedules and more or less formalized agreements, what is the status of those parents who formally have the custody and legal responsibility for children? In addition to these objective difficulties related to residential and legal arrangements, the identification of individual parents as being alone in assuming parental responsibility—financially, educationally, and emotionally—complicate the picture even further.

Given the importance of the timing and sequencing of events—both as independent and dependent variables—in demographic and social-policy analyses, it is important to understand how the transition to and out of lone parenthood is defined and experienced. The boundaries between couple-parenthood and lone-parenthood cohabitation may be much blurrier than is generally assumed. In this regard, loneparenthood endings may also be blurred, not unidirectional, and often involve ambivalence and gradual entry into a new relationship.

<sup>3</sup>This is different from the case of France, for instance, where two official definitions exist: one residential statistical definition and one legal administrative definition. The latter is historically related to the creation *allocation de parent isolé* (isolated parent benefit) introduced in 1976 (Letablier 2011), which assigned social benefits to eligible lone parents. Eligible lone parents are those whose children are not financially supported by the nonresidential parent. The residential definition alone was not appropriate to discriminate those households.

Our study is an ongoing project that aims at capturing both the diversity of pathways into lone parenthood and the subjective experiences of being a lone parent in Switzerland through a mixed-methods design. The quantitative part, based on longitudinal survey data, describes and analyses trajectories in and out of single parenthood as well as analysing specificities of lone parents in comparison to nonparents and cohabiting or married parents with similar characteristics (e.g., Struffolino et al. 2016). In the qualitative part of the project, we are interested in the objective adjustments one takes when entering the new lone-parent status, arrangements for the daily care for children, in working life, and in relational life. The subjective recognition of one's own state of lone parenthood takes time; it is often nonlinear and marked by ambiguity toward the role of main carer for the children and ambivalence in the acceptance of such a role. For this chapter, we restricted the analysis to the crucial aspect of defining the boundaries of the transition to lone parenthood and how parents define lone parenthood. In particular, we draw on a set of semistructured interviews with parents who identified themselves as "parents raising their children alone" at the moment of the interview in order to show the challenges in defining the beginning and the end of a lone-parent status. In lone parents' discourses about their transition to lone parenthood, objective residential arrangements, legal arrangements, and subjective feelings about their relationships and about the ways in which caring is shared are all important elements of the definition.

#### **Data and Methods**

We analysed 40 open-ended interviews reconstructing the life course of individuals who, at the moment we met them, were living as lone parents (i.e., men and women who identified themselves as parents raising their children alone). All the respondents lived in urban settings in the two Swiss cantons of Geneva and Vaud.4

We purposively built our sample to explore a variety of lone-parenthood experiences. In particular, we aimed at having variations in the patterns of entry into lone parenthood. One group of respondents entered lone parenthood as a consequence of having lost their partners, either through union breakup or a partner's death. In both cases, parenthood was experienced while in partnership and followed by separation, divorce, or widowhood. When the other parent was alive, we included cases in which parents were in contact with each other on a regular basis and cases where contacts were occasional or absent (either because of conflict or because of important geographical distance). The important aspect here is that parenthood began as a couples' experience, transformed itself through the separation and postseparation processes, and had to be unmade and rebuilt under a different configuration of relationships. A second pattern is represented by a transition to lone parenthood that began as a lone experience and not within a couple. These are cases of unplanned

<sup>4</sup>A short description of the main characteristics of the cases is included in our sample is in the Appendix.

children conceived because of contraceptive failure with an occasional partner, cases in which a partner has not accepted the role of a parent from the very start, or, more rarely, conceptions obtained through a donor. Often in these circumstances, the mother experiences lone parenthood beginning at the birth of her child. In other cases, the nonresident parent may have existed and known about and had occasional contact with the child. Parenthood then "belongs" to the individual from scratch and is constructed as an individual experience throughout.

The 40 respondents we interviewed were recruited through a multiple-entry snowballing technique. Initial seed individuals were reached through personal contacts, mailing lists from lone parents and family associations, researchers' contact boxes at a kindergarten, and flyer distribution in public places visited by parents and children. Entering the field was, as often happens, a gradual process. Flyers and mailing lists did not produce as many results before we entered contact with the family associations and before the word spread among lone parents' own contacts through snowballing. This technique has the advantage of reducing self-selection and reaching out to individuals who would not otherwise be willing to answer an anonymous phone call or e-mail from someone requesting participation in a survey. The fact that we gathered our initial contacts from several different sources reduces the biases that may occur when interviewing only a group of people in contact with each other, who could be similar to each other and therefore give access to only a specific experience of lone parenthood.

We also limited the interviews to parents who experienced a relatively recent transition to lone parenthood (mostly within 1–5 years), who had in most cases a child aged between 0 and 10 years,5 and who had full legal custody of their children as much as possible. These choices were made after a pilot study with larger inclusion criteria, which showed that recalling transitions happening further in the past was not easy, and lone parenthood when children are older and more autonomous is a completely different experience. In addition, the focus on recent transitions is justified by the fact that our longitudinal setup will enable us to follow the evolution of our lone parents in the medium run (up to 4 years from the previous interviews) and to understand the effects of duration on lone parenthood and the process of exiting lone parenthood. The focus on younger children is justified by the fact that these are children who still require a relatively high amount of care (i.e., they need continuous adult supervision). We did not include cases in which child custody was equally shared between the two parents, since coparenthood implies a different kind of relationship among parents and parents and children, and we wanted to limit our sample to lone-parenthood situations.

Interviews were extensive; on average they lasted 90 min. The sensitive nature of the topic caused emotionally intense experiences for both respondents and interviewers6 at times. Participants were asked to sign an informed-consent form, which sometimes generated concerns, even in cases in which parents happily agreed on

<sup>5</sup> In only two cases were children of preadolescent ages.

<sup>6</sup> In one case, after 2 h of interviewing, a respondent asked to withdraw from the study, citing that she could not bear the emotions. This respondent was taken out of the sample.

being interviewed. The interviewing team was composed of experienced researchers. The team met regularly to adjust the interview guidelines with the experience gained in the field and to discuss specific cases and situations that challenged our questions and methods.

The interview content began with questions designed to create sketches of the participants' life-course trajectories in different domains (union, family, education, and employment). We then asked participants to place parenthood and the transition to lone parenthood in the context of the life course. We probed for different topics: the evolution of the relationship with the other parent (when appropriate) and of his/ her relationship with the child/children; the relationship of the child/children with the respective parents' families; the current legal arrangements for the child/children; the possible employment adjustments related to lone parenthood; the role of institutional support in the transition and afterwards; the daily-life organization and childcare arrangements; the perceived advantages and disadvantages of lone parenthood; the partnership situation at the moment of the interview; and the current network configuration and social support.

We used a semistructured interview guideline. The openness also allowed for exploring the justification given for any behavioural patterns, meanings attributed to choices, and perceptions and expectations. All of these elements were crucial to understanding family-related processes described by statistical analyses. Each interview was tape recorded (after receiving participants' consent) and transcribed verbatim. The analysis of material was driven by the research question (What is the respondent's description of his or her transition to lone parenthood? What are the markers of the transition to lone parenthood he or she introduced as crucial in the process?). Even though the interview guide was the same, the themes emerged at different stages along the conversation. Analyses are interpretative and based on comparisons of individual cases and cross-case thematic coding. Top-down coding—based on the topic guide—and bottom-up coding—generated by spontaneous parts of the interview—were combined to produce categories that identified the salient characteristics defining the transitions to lone parenthood (Corbin and Strauss 1990).

#### **Findings: Marking the Transition to Lone Parenthood**

The main question of interest in this paper regarded how lone parents defined lone parenthood and what they identified as the marker or the markers of such a transition. The rationale was to ensure that we produced valid survey indicators to analyse transitions into and out of lone parenthood. Transition markers that we usually found in surveys may have diverged greatly from subjective markers of lone parenthood indicated as salient by the lone parents we interviewed. Under these circumstances, modelling the transition to lone parenthood in a valid way was an open challenge. In the next section, we report the way in which interviewees marked their transition to lone parenthood. We have three sets of cases: First, we have cases in which respondents identified a number of markers that corresponded to objective

and easily recordable events (end of cohabitation and legal separations for instance). Second, we have a large number of cases in which parents referred to changes in their lives that for them were pivotal in marking the onset of lone parenthood but were not often used to model the transition to lone parenthood (e.g., realising that one will have to care for the children alone, in-house separations, inner resolution to assume parenthood alone). The degree of correspondence between the way in which parental responsibilities are formally agreed on and the way in which they are arranged on a day-by-day basis varied substantially (e.g., parents with formal full custody and authority who actually lived in a situation of shared custody and did not feel they were lone parents and vice versa). Third, we examined cases in which the boundaries between couple parenthood and lone parenthood were blurred for the interviewee him or herself and in which individuals were ambivalent about their status as lone parents.

In what follows, we flesh out our argument by illustrating each of these situations with individual cases.

#### *Objective Markers Usually Used in Modelling Union and Family Transitions*

In a number of cases, residential and legal markers (residential separation, legal separation, decisions about the financial responsibilities for children, the children's custody, and authority) are sufficiently clear-cut to allow parents to refer to them as marking the start of lone parenthood. An objective marker of another nature is the diagnoses of pregnancy for women not in a relationship. Typically, interviewees identified a series of circumstances and described the transition toward lone parenthood as a gradual process composed of several events, making it clear that the transition to lone parenthood is multidimensional.

**Case No. 1** Antoinette (legal separation): "I knew already that there was something that did not work, but putting words to it ... I knew it did not work but I did not have the courage to leave..." She was a housewife or worked occasionally until she found a full-time job but left to go back to school. This accelerated the crisis of her marriage: "He did not accept the separation. He said to the judge that I should absolutely not work because it is impossible to study and raise the children; she must stop studying." The separation was formalized 1 year later, and it is this official date that Antoinette cites as the beginning of her lone parenthood (Antoinette 41, two children aged 13 and 16).

**Case No. 2** Céline (residential separation): Céline has been in a relationship and cohabiting with a man for about 6 years. She became the mother of two children but after the second child, she definitely realized that the relationship with her partner was gradually deteriorating. 2 years after the birth of their second child and after several attempts to make her relationship work again, she decided to break up with her partner. Her residential separation from the partner marked the watershed in this story: "I think the birth of the kids did not ... did not help our relationship ... We wanted to have children, but it's true that afterwards there was no more feeling in our relationship ... Sometime after the birth of the second ... we were parents, but as to our relationship we were living as ... I don't know, as friends ... we could perfectly deal with everyday family life but there was nothing more in our life as a couple ... We tried to find the means ... we did not ask for help, but we have tried ... but we did not succeed ... and then we decided to separate in October 2010 ... At the beginning, I wanted to keep the flat, it was a four room flat, and he would find a smaller one, but then I found [one] by chance." Since then, Céline has had full custody of her children and receives regular payments from her expartner: "Yes, he gives me what we stated in the partnership agreement, then he takes them one evening every week and once every two weekends ..." (Céline 34, two children aged 4 and 6).

#### *Relational and Subjective Markers of the Transition to Lone Parenthood*

Some of our stories speak about lone parenthood as starting with in-house separation, in which the parents still live together but one of the two clearly withdraws from his or her role both as a parent and as a partner. Here, the timing is more diffused. It refers to relational changes rather than factual events: the other parent begins a new relationship, the worsening of conflicts or the interruption of contacts between parents, and a growing inner resolution to assume parenthood alone.

In Marie-Jo's account and, more dramatically, in Arthur's story, (both of their cases are described below) residential and legal objective markers occur much later than the beginning of their lone parent status. In such situations, the relationship as a couple is nonexistent and only one parent actually cares for the children, despite the fact that the parents continue to cohabit. These couples would be coresidents and even married in the case of Arthur, and they officially share responsibilities and rights for the children.

**Case No. 3** Marie-Jo (in-house separation): Marie-Jo was in a relationship with a man whom she began to live with soon after they met. Her partner convinced her to have a child together, but once she was pregnant, he seemed to have lost interest in parenthood. This became evident when she entered her sixth month of pregnancy. According to Marie-Jo, "The situation became very difficult at home ... I think I have lived through this pregnancy alone ... We were living under the same roof, but we would not meet each other. He started sleeping in the other room. He would go to work early in the morning, he would come home late ... I mean [we were] not even housemates." The situation remained the same until the child was 6 months old: "He (the partner) would come home in the evening ... he would not even hold the baby in his arms ... he would hardly look at her." Marie-Jo dates the beginning of her lone parenthood from a long time before the residential separation, which happened when the baby was 6 months old. To her, the transition started when the father of the child began withdrawing from his role as a partner and as a father simultaneously. She experienced his leaving the common household as the inevitable consequence of her lone parenthood, not its onset. Despite the father financially contributing the sum of money legally established, he has not seen his daughter since the separation. Marie-Jo feels that she was the only parent caring for her daughter since her birth. All in all, Marie-Jo's residential trajectory does not overlap with her trajectory as a lone parent (Marie-Jo, 37, one child, aged 5).

**Case No. 4** Arthur (in-house separation): Arthur has the custody of his two daughters of 5 and 3 (they were 2 and 4 when the parents wanted to break up). When his wife left the apartment, he started filling out the paper work for a formal separation, which was not ready when she decided to move in again after 6 months: "The lawyer said she has the right to come back home. So she came back but meanwhile she had started a new life. I have to say that I was alone to care for the children. There was absolutely no love anymore. So it was extremely difficult when she came back, just impossible to bear it…. It was difficult in relation to the children, because I prepared them, I told them, I explained to them, but the fact that they saw her coming back, it was difficult, in the sense that they thought 'ah, here they come back together.'" Arthur's wife used to bring her temporary relationships to their home, leaving Arthur to care for the children as if he were alone. The tension between them rose to the point that she tried to stab Arthur in front of the children. Only at that point could legal measures be taken to oblige the mother to leave the common household (Arthur, 31, two children aged 5 and 3). In such situations, it may happen that legal and residential markers of lone parenthood do not correspond to the parents' daily arrangements or their understandings of the situation.

**Case No. 5** Anouk (residential separation but formally shared parenting): Anouk's lone parenthood status was hidden by the formal shared custody. Anouk separated from the father of her child in 2011 due to his alcohol addiction. She simply moved out. She does not receive any alimony and there is no legal arrangement about the father's visiting rights given that what was legally foreseen in the case of separation was shared custody and authority. In reality, the father does not care for the child regularly. Social services workers know that Anouk has troubles claiming maintenance payments from her partner due to his alcohol addiction, but since the father has never abused the child they refuse to mediate. The only way for Anouk to have her rights recognized would be to go to court and ask for a change in the legal arrangements, which she is not ready to do since she is still very much emotionally connected to the father of the child. She prefers to find her own way to regulate parental responsibilities, independently of what the laws say: "So I said, 'okay, you can have your son every week, one day in the weekend, and one day on weekdays, there is no problem, when you want,' but he had to show me that he was sober. So we set up a system whereby I would make him take a breathalizer test. I had ... a device, and so I was a controller" (Anouk, 41, 1 child, aged 4).

**Case No. 6** Susan (residential separation but formally shared parenting): The case of Susan illustrates a similar incongruence between legal and actual arrangements. Subjectively, the transition to lone parenthood for Susan started with the resolution to leave the country of common residence and return to Switzerland. However, no objective markers of transition can be tracked down, as she is still formally married according to African law and has no legally registered children according to Swiss law. Susan has lived as a lone parent in her daily life for a year, but the separation from her husband has not yet been formalized. She worked and lived for 10 years in Africa where she met the father of her children, who still works seasonally there. After a five-year relationship, they had the first unplanned child and then a second intended one, as a couple. Susan felt that their relationship was not working anymore after this second birth and returned to Switzerland with the children to have her mother's help with childcare. Despite being de facto separated for a few years at the moment of the interview, she has not started any formal separation from her husband yet: "So we did not start any formal procedure to end the separation process, and even if we did, we know that it would be complicated ... to arrange something regular. Because ... he has a job that is not regular." In addition, due to administrative complications, her children do not have official documents either as foreigners or as Swiss nationals. This is despite the fact that they attend school regularly as residents of a Swiss territory; otherwise, they are not officially acknowledged by the Swiss government (Susan, 36, two children, aged 7 and 5).

**Case No. 7** Catalina (flexible custody arrangements): In contrast to Anouk's and Susan's cases, there are situations in which lone parenthood is formally recognized but parental practices imply day-by-day shared parental arrangements. Catalina, a mother of a six-year-old child, is actually closer to living in a shared custody situation than being a lone parent. Catalina separated from her cohabitant partner when her daughter was 6 months old and has formally obtained full custody and authority over her child. She said, "He has the right to visit her ... every Thursday evening ... and one [out of] every two weekends." However, when reporting on her day-to-day parental practices, we find out that on many occasions, the formal custody agreement is completely reversed. She said, "It was good for me that he was not working in a way ... because last year, he took her during the whole school break.... Yes, the whole week, and I would stay with her over the weekends" (Catalina, 40, one child, aged 6).

#### *Ambivalence About Lone-Parenthood Status*

The previous cases point out two important findings: first, there are multiple markers of the transition to lone parenthood that are relevant for parents depending on their parental and relational history; and second, objective markers do not always correspond to the onset of the experience of lone parenting. There is one more set of answers that could not be coded within the two previous groups of cases. For some respondents, it is not possible to univocally define their status as a lone parent or a parent in a couple. On the one hand, they answered our call for interviews with lone parents, but on the other hand, they attached a strong meaning to their relationship with the other parent, who was still perceived as a partner. As a result, they expressed a considerable amount of ambivalence about their own parental status.

**Case No. 8** Alexandra (ambivalent relationship with the child's father): After her divorce, Alexandra met Edouard, a professional who was living with his wife and children. Alexandra moved to be closer to him and continued being his lover, regularly spending holidays with him. Their 13-year secret relationship ended when she became pregnant. She kept the child despite the father initially trying to convince her to have an abortion and refusing to recognize him or see him. After 2 years of separation, Alexandra and Eduard started seeing each other again, and the father started to spend time with his child, introducing him to his official family (Alexandra, on the contrary, did not have access to them until now). She said, "From my side, I have always considered him as my partner and introduced him as such to my family, to my friends, and to others; in his circle, it is rather the opposite … I stay illegitimate, forbidden, etc." Alexandra lives as a lone parent in her daily life, but at the same time, she is in a sort of living-apart-together relationship with the father of her child and is conscious of the inner contradiction in her situation. "Again, there is a rather fundamental contradiction given that it is not long that I felt ready to live with Edouard, not like a fusion-like couple, but like we had already discussed for years ... we would need a duplex or an apartment with two entries so that we can be separate and together when we want" (Alexandra, 45, one child, aged 3).

#### **Discussion**

Lone parenthood is a form of "doing and being" a family. It is sometimes a transitory period in family development and sometimes a long-term experience of parenting alone. This paper draws on qualitative interviews belonging to an ongoing project on lone parenthood in Switzerland. The aim of the paper was to study the ways in which men and women who are raising their children alone narrate their transition to lone parenthood and when they identify the onset of their lone parenthood. By focusing on relational configurations and practices, rather than on economic and residential living arrangements only, this study calls into question the usual assumptions on which most studies on lone parents are based.

We point out at least two limitations of the current practices in quantitative research. First, we show the limitation of residential and administrative criteria to delimit lone-parent status and which are used as bases for current statistics. Lone parents are most often defined as parents residing with at least one child in the absence of the other parent. Variations in such a definition concern the presence of other adults in the household, like a new partner (stepfamilies), other relatives or unrelated adults (multiple generation or enlarged families), or the maximal age of the dependent children present in the household (16, 18, or 25). Alternatively, lone parents are those parents who are financially and legally responsible for the children alone. Such objective definitions are most useful for counting aggregate population data and for having a gross estimation about who is entitled to receive child allowances, child support, and social support. Yet, were our interest as family scientists to be the measuring and modelling of individual lone-parenthood trajectories and understanding their implications as life-course experiences, then we would need to

consider subjective markers such as those pointed out by our interviewees in addition to more standard objective indicators. Family surveys, particularly longitudinal perspective surveys, would ideally include measures of social isolation, asking for instance whether the respondent feels to be the only person responsible for their children's development. Analyses of lone parents' financial situation would assay respondents to capture retrospectively whether their financial hardships had begun before or after lone parenthood even in cases when individual incomes were unchanged or had increased. Similarly, one could monitor variations in relational quality or in the amount of time that each parent spends with the children. Relational criteria are as important as administrative or residential criteria for identifying whom policies should be addressed to, but they are also crucial for capturing the meaning and effects of lone parenthood in the life course.

Second, our findings call into question the assumption that parental and partnership trajectories evolve simultaneously in the transition to lone parenthood. In fact these trajectories do not always follow the same timing (e.g., Anouk's case) and this may cause a mismatch in the way lone parenthood is measured and modelled in qualitative and quantitative research. Our interviews show that many current lone parents who were in a couple take some time to end the relationship with the other parent. Cohabitation may not be coincident with a partnership and with coparenting (e.g., in Arthur's case). These findings suggest the importance of looking at how families "do family" through the meanings they attach to relationships. However, it may also be difficult to describe the status of a relationship. Those who were not in a relationship when a child was born can be ambivalent about their partnership status in relation to the other parent (e.g., in Alexandra's case). The transition to lone parenthood, particularly lone parenthood as a consequence of separation or divorce (as analysed in this chapter), is often a gradual and ambivalent transition, which involves a variety of dimensions. The fact that it is gradual makes it hard for respondents to identify a date at which it began. The ambivalence makes it hard for them to identify the relevant marker for the transition. In addition, it is not necessarily a unidirectional transition since separation and reconciliations happen without being marked by consequent legal and residential steps. The latter characteristic increases the risk that retrospective data misestimate periods of lone parenthood. The gradual evolution from couple to lone parenthood or from a pregnancy with a potential future partner and coparent to lone parenthood includes some analytic challenges. Demographers have often felt the need to assume that there is a clear distinction between being single and being in a partnership. This is even truer in the case of a pregnancy or a child. However, as we have seen, this is not always the case. What are the consequences and why do we care? For instance, if we were to study lone parenthood together with noncustodial parenthood, we may find mismatches in the declarations of men and women concerning the start of their respective new situations (some may think of themselves as in a partnership and some as already single). These challenges are not unique to lone parenthood and are strictly related to similarly blurred frontiers and definitions of cohabitation and Living Apart Together relationships (Binstock and Thornton 2003; Manning and Smock 2005). By pointing out the complexities associated with studying lone parenthood, we hope to contribute to the field by informing future data collection and empirical analysis of lone parenthood.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **Annex 1: Respondents' Profiles at Interview**


a Lone parent

b One lone mother was pregnant at the time of the interview

c Low level of education includes participants without any university degree (including VET degrees). High level of education includes participants with a least a bachelor degree

#### **References**


**Laura Bernardi** is Professor of Life Course Demography and Sociology at the Faculty of Social and Political Sciences at the University of Lausanne, Switzerland, and Deputy Director of the Swiss National Center for Competence in Research LIVES. Her major research areas are fertility, family demography, migration and life course. She has published in major sociological and demographic journals on topics like fertility intentions and behaviour, social network influences on family development, intergenerational transmission of fertility and family formation, lone parenthood and health, migration and family formation. She is also co-editing the Springer series on Life Course and Social Policies.

**Ornella Larenza** is a Social Sciences PhD student at the University of Lausanne. She works at the National Centre for Competence in Research LIVES at the University of Lausanne where she is involved in a qualitative panel research on lone parenthood in French speaking Switzerland. Her main areas of research are family sociology and policy, care, social policy and life course.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Part II Demographics of Lone Parents**

## **Chapter 6 Are Lone Mothers Also Lonely Mothers? Social Networks of Unemployed Lone Mothers in Eastern Germany**

**Sylvia Keim**

In Germany as in many other western countries, lone parenthood is a common family form: a single parent heads every fifth family with children under age 18 (BMAS 2013: 5). There are, however, considerable differences between western and eastern Germany: in 2009, 17% of families in western Germany were headed by a lone parent, compared to 27% in eastern Germany (Lois and Kopp 2011). Around 90% of lone parent families are headed by a mother (BMAS 2013: 5), and most research on lone parents is on lone mothers. Lone parent families are a very heterogeneous group in terms of their socioeconomic characteristics, their path into lone parenthood, and the ways they perceive and cope with lone parenthood. Regardless of these differences, all lone parents face competing pressures of holding down a job while caring for their children. Because combining family life and work is challenging, especially without the support of a partner, lone parents are often unemployed. Currently in Germany, 40% of lone parent households are receiving transfer payments for the unemployed ("Grundsicherung für Arbeitssuchende") (BMAS 2013: 5), and about half of them have been receiving these payments for more than 4 years (Achatz et al. 2013: 12). Compared to couple families, lone parent families are more likely to live in poverty, and to report having high levels of stress and low levels of well-being (BMAS 2013). Lone parents and the long-term-unemployed are often designated as vulnerable groups. The concept of vulnerability describes a zone in which the risk of social downward mobility and poverty is high, and an individual's abilities and resources to cope with these social risks are limited (Vogel 2006). When a woman who is a lone parent and/or is long-term unemployed is faced with critical life events, chronic stresses, or environmental hazards, her resources are often not sufficient to buffer the strains; thus, she becomes increasingly vulnerable (Hanappi et al. 2015), and her disadvantages accumulate.

S. Keim (\*)

© The Author(s) 2018 111

Institute of Sociology and Demography, University of Rostock, Rostock, Germany e-mail: sylvia.keim@uni-rostock.de

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_6

In recent decades, research on social support has shown that having supportive personal relationships can ease or even help individuals overcome situations of poverty, unemployment, or stress; and can contribute to well-being (Cobb 1976; Henly et al. 2005; Lin et al. 1999;). A number of studies have shown that social support benefits single parents in particular (Ciabattari 2007; Cook 2012; Harknett 2006). While social support research has mainly focused on the perception and the reception of different types of support, social network research takes a more differentiated view by looking at the meaning and the effects of social relations, and the structural properties of the networks they form (Smith and Christakis 2008). This structural perspective examines the structural characteristics of the network (such as size, density, and composition) and the social mechanisms beyond those of social support (for an overview, see Berkman and Glass 2000), and analyzes social relations and network structures separately from the mechanisms through which social relations affect individual behavior and well-being. Thus, this perspective enables us to determine which types of ties are supportive in which ways, as well as how the overall network structure influences the perception and the reception of support.

Little is currently known about the interplay of social network structures and social support in creating well-being or in enabling people to escape poverty. To help fill this gap, we focus in this article on long-term unemployed lone mothers who have experienced two critical events—namely, separation from a partner and the loss of employment—and who could therefore be perceived as being especially vulnerable.

Studies on lone parents and on the effects of unemployment have often postulated that the critical events of separation or job loss lead to social isolation, or at least to a reduction in social contacts. It is therefore assumed that a woman who is separated or unemployed loses certain ties, such as ties to her ex-husband, his relatives and friends, or her former colleagues. Additionally, people who lose a partner or a job also lose important sources of social approval (e.g., Castel 2000; Myers et al.1975). However, only a few existing studies have focused in detail on the personal relationships of members of these vulnerable groups, analyzing the social networks they form, the structure of these networks, their interactions, and how they are perceived. These studies have examined not only what these individuals have lost, but also how their lives have changed, and how they have actively built their social networks according to their preferences and needs. The results of this research indicate that people who are unemployed or lone parents have a large variety of networks that differ in terms of structure and composition, but that they are rarely fully isolated individuals. Additionally, these studies have found that individuals who have lost certain ties tend to compensate for these losses by building new ties or intensifying existing ones (e.g., Marquardsen 2012 for unemployed persons; Niepel 1994 for lone mothers).

In order to provide a more detailed picture of the social relations of unemployed lone mothers, we present in this chapter an explorative study in which we examine how the networks of unemployed lone mothers are structured and composed (e.g., whether these women can be described as socially isolated), identify different types of networks, and analyze which types of networks are or are not supportive and conducive to the well-being of women in vulnerable situations.

#### **Theoretical Background: Subjective Well-Being and Social Networks**

As our main interest is in measuring the quality of life of unemployed lone parents, we focus on their subjective well-being. Researchers have not agreed on a clear-cut definition of the concept of well-being; a number of scholars have alternatively employed concepts such as happiness or life satisfaction, and have sometimes even used these terms interchangeably (e.g., Ferrer-i-Carbonell 2005, Kohler et al. 2005; Myrskylä and Margolis 2014). However, many researchers, especially those in the socio-psychological tradition, have characterized subjective well-being as a broad concept that encompasses emotional elements (positive and negative emotions, such as joy, shame, happiness, depression, contentment, anxiety, affection, and stress) and cognitive elements of life satisfaction in general, and of satisfaction in different domains (e.g., work, family, leisure, health, or finances). Scholars have also described subjective well-being as a concept that encompasses past experiences, future expectations, and the views of significant others (Diener 2007; Heidl et al. 2012; Kim-Prieto et al. 2005; Glatzer 1992). The effects of social engagement and social relations on subjective well-being are frequently stressed (e.g., Cummins 1996; Diener and Seligman 2002, Helliwell and Putnam 2004).

In our research on unemployed lone mothers' subjective well-being, we take the perspective of social network analysis. The social network perspective not only stresses the relevance of the existence of ties and the support they provide, but also considers the structure of the social relationships the individual is embedded in. The properties of the network partners (alteri), as well as the properties of the network structure, enable or constrain an individual's (ego) behavior (Burt, 1982; Degenne and Forsé, 1999). How social support affects an individual's well-being depends on the structure of the network and on the network partners' resources. The concept of social support refers to various kinds of perceived and received support, with distinctions often being made between emotional support (providing affection, respect, and advice), social companionship (forms of socializing, such as meeting and going out), instrumental support (helping with chores or child care), and information (e.g., providing information about job opportunities, Granovetter 1973) (cf. Diewald 1991; Hirsch 1980; Wellman and Wortley 1989). In particular, emotional support and social companionship foster the affectual aspects of individual well-being (by, for example, reinforcing identity and recognition; as Lin 2001 explained), while instrumental and informational support pertain more to the cognitive aspects of life satisfaction.

From a network perspective, the overall level of support and the specific types of support available to individuals vary. The types of alters who provide support, and the resources these alters have available also vary (Wellman and Wortley 1989). In addition, network characteristics differ depending on the ego's social position (e.g., socioeconomic status, marital status, number and age of children). These differences can make it difficult to compare the results of different groups (McLanahan et al.1981; Acock and Hurlbert 1990, 1993).

Of the many network characteristics that could be analyzed, I will present here four of the characteristics that are most commonly used in research: network size, composition, density, and multiplexity.

The size of the support network measures the number of individuals who provide support, while the composition of the network refers to the characteristics of the network partners, including their role in relation to the respondent. Research on the associations between network size, support, and well-being does not show consistent results: on the one hand large networks are associated with happiness and wellbeing, regardless of the exchange content (Gerstel et al. 1985); while on the other mothers in smaller networks tend to receive more support than those in larger networks, and to evaluate the support in a more positive way (Malo 1994). Research on network composition has shown that having a variety of types of relationships—that is, having a heterogeneous network—increases self-esteem, and may therefore improve well-being and access to instrumental resources (Acock and Hurlbert 1990). Networks that include a partner, relatives, and friends or neighbors significantly contribute to life satisfaction (Deindl 2005).

Other relevant network measures are the density and the multiplexity of ties. Density measures the degree to which network partners interact with each other. Multiplexity is a quality of the relationships in the network: multiplex relationships provide various forms of support, while uniplex relationships only provide one type of support. In general, it is assumed that dense and multiplex networks are positively associated with providing services and emotional support, while loose-knit networks are positively associated with providing access to information and to new social contacts (Mitchell 1969; Granovetter 1973). However, research has also shown that this might not be the case for lone mothers, because they may see support from relatives, usually within a dense network, as less satisfying than support from friends within a sparse network (McLanahan et al. 1981); and because less dense networks may provide more flexible support, and are therefore better at providing the support needed when adapting to a new situation (Gerstel et al. 1985).

Social support cannot be demanded or granted arbitrarily; rather, these exchanges follow specific culturally defined rules and practices that define the expectations, obligations, and rights of the individuals involved (McCallister and Fischer 1978). One concept that is relevant in this context is reciprocity. According to social equity theory, individuals in inequitable relationships will try to restore equity by providing support to their supporters (Walster et al. 1978; Plickert et al. 2007). However, there are also role relations with rather one-sided exchanges; e.g., parents supporting their (young) children. If we take into account the role of cultural norms and values in social interactions, we can assume that if there are regional differences in these norms, we will also find regional differences in the associations between networks, support, and well-being.

Support affects individual behavior and well-being through two mechanisms: 1) the direct effect of social support that enhances personal well-being in general, irrespective of the stress level; and 2) the buffering effect of social support in the presence of a stressful event (Cohen and Syme 1985). In our research, the respondents had experienced at least two potentially stressful events: separation from a partner and unemployment.

Personal relationships are not, however, necessarily supportive or conducive to well-being, as they can also restrict behavior, have negative consequences, and be sources of conflict (Abbey et al. 2010; Attree 2005; Gräbe 1991; Acock and Hurlbert 1990; Laireiter et al. 2007). Little is known about the interplay between support and conflict. Recent research has shown that among respondents who reported having a high level of social support social conflicts were uncorrelated with their well-being, while among respondents who reported having a low level of support social conflicts had a negative impact on their well-being (Abbey et al. 2010).

#### **Lone Mothers' Well-Being, Social Networks, and Support**

Research on subjective well-being and happiness has shown across countries that parents who are in a partnership have higher levels of well-being than single parents, irrespective of the specific definition of well-being (BMFSFJ 2012; Helfferich et al. 2003; Kohler et al. 2005; Osborne et al. 2012; Vignoli et al. 2014; Winkelmann and Winkelmann 1998). But what are the mechanisms that underlie the correlations between having a partner and well-being? The low levels of well-being observed among lone mothers are often attributed to the lack of a stable partnership, the stress and responsibility associated with childrearing, and the difficulties associated with finding employment. While research on partnerships has clearly shown that all people who are in a partnership (irrespective of their parity) have higher levels of wellbeing than people who are single (Kohler et al. 2005), and research on unemployment has shown that people who are unemployed have lower levels of well-being than people who are employed (Clark and Oswald 1994, Winkelmann and Winkelmann 1998), research on the correlation between having children and well-being has generated mixed results (Kohler et al. 2005; Myrskylä and Margolis 2014). If we look at lone parents specifically, it appears that their relatively weak socioeconomic position largely explains their lower levels of well-being (Vignoli et al. 2014).

Many studies use cross-sectional data, which cannot address causal relationships. Do lone parents report lower levels of well-being than parents in partnerships because of the hardships of lone parenthood, or are people who are in general less satisfied with their life also at higher risk of becoming a single parent or unemployed? The longitudinal data that would be needed to address these questions are sparse. One study on lone mothers in Poland that controlled for selection effects found that the birth of a non-marital child does not decrease women's well-being, and may have the opposite effect (Baranowska-Rataj et al. 2014). Similarly, a study in Great Britain uncovered strong selection effects; specifically, the study found that compared to their middle-class counterparts, women from working-class backgrounds are more likely to have non-marital children, and they are less likely to be employed or to repartner (Rowlingson and McKay 2005). In sum, previous research indicates that both selection effects and the effects of the individual woman's socioeconomic background and partnership status influence her subjective well-being. However, more longitudinal studies will be needed to disentangle the role of these effects and the mechanisms involved.

How do social relationships, social networks, and social support affect lone parents' well-being? Support research has shown that compared to coupled parents, lone parents receive less support and are more likely to report a lack in support (Targosz et al. 2003). However, single parents strongly rely on social support (Balaji et al. 2007; Harknett 2006), as having support enables them to combine family and work (Cook 2012; Ciabattari 2007). Having social support has been shown to be associated with enhanced economic well-being (Henly et al. 2005), while having access to child care and having emotional support has been found to be positively linked to the quality of mother-child interactions (Belle et al. 1981). By contrast, having low levels of support has been shown to be associated with depressive symptoms and low self-efficacy (Harknett 2006). A strong degree of reliance on support has been found among low-income single mothers in particular, and receiving support has both positive and negative consequences (Barnes 2003). Support from friends seems to be especially important (Coletta 1979; Tietjen 1985), but not all women have access to friends (Balaji et al. 2006). Support research has mainly focused on the positive effects of social relations, while the negative effects of social support (Attree 2005; Gräbe 1991) and the effects of negative ties and conflictual relations (Acock and Hurlbert 1990; Laireiter et al. 2007) are rarely discussed.

The network perspective promises to provide a more comprehensive view of social relations, integrating the different kinds of ties, the different mechanisms through which social relations affect individual behavior, and the network structure. However, the literature on single parents from a network perspective is sparse. Most of the few existing studies were conducted in the U.S. and focused on women who are divorced (Albeck and Kaydar 2002; Duffy 1993; Leslie and Grady 1985; Nelson 1995), rather than never married. The research conducted in the U.S. on the social networks of individuals with different marital statuses has shown that the networks of individuals differ considerably depending on whether they are divorced, never married, or widowed (Hurlbert and Acock 1990). Little is known about how the networks of single parents living in Germany today differ depending on their paths into lone parenthood. One exception is a study by Nestmann and Stiehler (1998) on parents in Saxony (an eastern German federal state in which the share of mothers who are never married is relatively high), which found that never married mothers differ from divorced mothers in terms of network composition and access to support. We therefore assume that the results from research on divorced women in the U.S. cannot be fully applied to Germany, where over one-third of lone parents have never been married, 42% are divorced, 17% are not formally divorced but live in separate households, and 6% are widowed (Statistisches Bundesamt 2010: 11). We also have to keep in mind the welfare system and the prevailing norms and values

regarding lone parenthood and giving/receiving support differ by region. For example, in contrast to the findings from the U.S., the results of a study from Sweden found that lone mothers receive more support than married mothers, and that the networks of married and unmarried mothers are very similar (Tietjen 1985).

Studies that focus explicitly on the interplay of social networks and social support for lone mothers, and their effects on personal well-being, are rare. Particularly interesting are the studies by McLanahan et al. (1981) on divorced mothers in the U.S. and by Niepel (1996) on lone mothers in Germany. Despite their different samples, the two studies can be compared because they both focused on network structure and composition, collected network data in a similar manner, and developed a network typology.1 Based on their interviews with 45 divorced mothers in the U.S. (all of whom were white and almost all of whom were employed), McLanahan, Wedemeyer, and Adelberg (1981) identified three types of networks that emerged following divorce: (1) the woman reunited with her family of origin, (2) the woman built an extended network, or (3) the woman maintained or reestablished a conjugal relationship with the ex-partner or a new partner. The first type of network was dominated by members of the woman's family of origin, included few friends, and was relatively small and dense. The second type of network included various groups of ties, was relatively large, was composed of several cliques, and included a large share of lone parents. In the third type of network, the new partnership dominated, and included the partner's relatives. This type of network could be either small and dense with many old friends, or large and sparse with many friends. All three types were found to foster individual well-being if they fit the mother's role orientation (aiming for stability in the first and third type, or aiming for change in the second type).

In Germany, Niepel (1994) explored the social networks of 20 lone mothers (most of whom were employed) in the city of Bielefeld. She identified three network types based on their composition of family and friends: (1) friendship networks that were composed mainly of friends, included no relatives, and were relatively small and sparse; (2) family and friendship networks that were composed of friends and relatives, were relatively large, and included many relationships of long duration that were either loose or dense; and (3) family networks that were composed mainly of members of the woman's family of origin, and were relatively small and dense. Niepel did not discuss how each type affects well-being, but

<sup>1</sup>McLanahan et al. (1981) performed in-depth interviews in which they let the respondents write their supportive ties on small disks and arrange them on a piece of paper, drawing lines between those who knew and interacted with each other. They used probes to ensure that the respondents considered all three theoretically predefined areas of support. In contrast, Niepel (1994) used within a qualitative interview on lone parenthood a questionnaire that in a first step asked about the relevance of 17 different forms of support, and in a second step demanded that the respondents provide a list of supportive persons. Despite their different methods of collecting the data, they both applied a very broad network-generating question, and were trying to collect a "full range of network as well as support characteristics" (McLanahan et al. 1981: 603).

observed that heterogeneous networks of medium size appear to provide the best conditions for coping with lone parenthood.

If we compare the network types of Niepel with those of McLanahan, Wedemeer, and Adelberg (1981), we can see that both typologies focused on whether a given network is composed mainly of friendship ties or of family ties. The network typologies have a few similarities and large differences. Niepel's family and friendship networks are similar to extended networks, but they are not necessarily large, nor do they contain a high share of lone parents. Niepel's family network resembles the "reunited with the family of origin" type, but she found only one woman who had this type of network. Moreover, Niepel's friendship networks do not resemble any of the types identified by McLanahan, Wedemeer, and Adelberg. On the other hand, she did not find the conjugal form in her data. These discrepancies could be attributable to cultural differences between the U.S. and Germany, historical differences between 1984 and 1996 (increasing individualization and the loosening of traditional family ties), or differences in the sample (only divorced mothers versus all kinds of lone mothers). Our study is in turn very different from these previous two studies in terms of the sample (all unemployed), the location (eastern Germany), and the time frame (conducted almost 20 years after Niepel's study). We therefore cannot expect to find similar types, but we will also focus on network composition.

In addition, based on our theoretical reflections we are also interested in including stressful relationships in order to capture the negative as well as the positive effects of social contacts on well-being.

Our analysis presented here focuses mainly on three research questions: (1) What is the structure and the composition of unemployed lone parents' networks? (2) Can we distinguish between different types of networks? And, (3) do these network types contribute to well-being in different ways, or do they detract from well-being?

#### **Methods**

Given the explorative nature of the research questions, this research uses a concurrent mixed-methods approach that allows us to link network structures with the narrations of processes and the related meanings (Onwuegbuzie and Collins 2007). We combined problem-centered interviews with a structured collection of social network data on the same set of respondents. The sampling strategy has to balance the breadth and depth of the research interest. Aiming at an extensive collection of detailed insights into the respondents' living situation and their social relations, asks for a small sample. Our sample therefore does not aim at representativity, but rather was designed to analyze in-depth the diversity of a certain population group that can be described as vulnerable. We purposively chose a group of persons, which is homogenous in being lone mothers, unemployed, and enrolled in public programs for unemployed parents in the federal state of Mecklenburg-Western Pomerania, Germany. Within this group, we aimed at a maximal variation, including persons with different age, number of children, children's age, duration of unemployment, duration of lone motherhood, urban or rural living situation. We realized 26 interviews. They took place from July to September 2013, and lasted between 1 and 5 h.

The problem-centered interviews included narrative incentives designed to elicit rich and detailed accounts from the respondents, and prepared questions that cover a range of topics strongly focused on our research interest (Witzel and Reiter 2012). Each interview started with an initial narration, induced by asking the respondent to explain what her life was like when her first child was born, and how her life developed from that point onward. The specific topics touched upon later on included: experiences with family life, work experiences, dealing with unemployment, balancing work and family, and future aims and social relations. Collecting retrospective narrations allows us to examine at least partially the subjective causality link between lone motherhood, social relations, and well-being. The emotional elements of the concept of well-being were covered by asking the respondent about her feelings regarding, for example, being a lone mother or being unemployed; while the cognitive elements were covered by asking the respondent to evaluate her life satisfaction in general and in specific domains. The retrospective interviews cover past and present experiences, and conclude with questions about the respondent's expectations for the future.

Toward the end of the problem-centered interview network data was collected with the help of VennMaker software. This tool allows for a structured collection of network data, but because it was introduced at the end of the interviews, respondents continued to talk and provide more detail about their social relations while interacting with it. The name generating questions asked the respondents to identify the individuals who provide them with emotional support (talking about personal matters), childcare support, and support in finding a new job; as well as the individuals with whom they socialize (meeting up, going out, eating together, exercising, etc.). The program also asked the respondents to identify the individuals to whom they provide support, and with whom they have conflicts. The respondents could then place the generated names into a map of concentric circles (similar to those used by Kahn and Antonucci 1980 and in our research on family networks, see Keim 2011). The characteristics of the reported individuals were then collected (e.g., gender, age, employment status, family status, and relationship to ego), as well as the contacts between the alters ("know each other well," "know each other less well," "don't know each other").

The qualitative accounts were recorded, transcribed, and analyzed using the qualitative content analysis method (Mayring 2000) by a team of researchers who identified the relevant themes and categorized the material based on how social relationships are perceived, activated, and generated; and whether they contributed to or detracted from the well-being of the unemployed lone parents and their children. Well-Being was coded according to its emotional and cognitive aspects: e.g., feelings about unemployment and lone parenthood, contentment with the family situation, and future expectations regarding family life and career. The network data were analyzed by means of social network analysis. We focused on network size

(the number of network partners), network density (calculating the actual ties between alters over all of the possible ties among alters), and the multiplexity of ties (being named from one or several name generating question), as they can be interpreted as measures of the degree of social isolation or involvement. We also analyzed network composition in terms of role relations and other alteri characteristics. We analyzed these measures separately for the network of support (built by the name generating questions 1–4). This analysis cumulated in the development of a typology of social network composition that—despite the small sample—generated four clearly distinguishable types. In a third step, the qualitative and quantitative analyses were brought together through comparisons of the narrations about well-being, support, and conflict with the networks structure and compositions condensed into four network types.

The respondents were recruited from public programs for unemployed parents designed to help them stabilize their family life and prepare for vocational integration. We interviewed individuals from different urban and rural areas of Mecklenburg-Western Pomerania. The interviews took place in July to October 2013 at a location convenient for the respondent. Each interview took an average of 2 h.

Included in this analysis are 26 mothers who were living alone with their children. The mothers were aged 24 to 49 (mean: 33.3). Most of these women had one or two children, and in most cases their younger child was between the ages of one and 15. None of the women we interviewed was freshly separated; all had been living alone with their children for at least 1 year, and some had never lived with the father of their children. Almost half of the women were in a new partnership, but they were not sharing a household with their partner. All of the women had lower secondary education or upper secondary vocational education. About half of them had never been in stable employment since finishing school or job training. Some of the women had been unemployed for more than 20 years, while four women had been unemployed for less than 2 years.

#### **Results**

#### *The Qualitative Accounts on the Current Life Situation and Well-Being*

We have identified five main themes related to well-being that recurred in the interviews: (1) general life satisfaction, (2) family life, (3) unemployment, (4) financial situation, and (5) social relations.

(1) General life satisfaction: Most of the respondents reported being satisfied with their life in general. However, many of them described having experienced severe strains (e.g., chronic or acute diseases, disabilities of the respondent or her children, exposure to violence, or large debts) or an accumulation of strains. Having experienced strains was not automatically associated with reduced overall well-being; rather, the extent to which the overall well-being of a respondent was negatively affected by these challenges depended on how she perceived them and was able to cope with them. In turn, the respondents who reported having low levels of life satisfaction were not necessarily those who suffered from multiple or especially severe strains.


the only jobs available to them were those that would require them to work evenings, nights, or weekends, but they had no one to care for their children at these times: "I go in circles for years." In contrast, the respondents who said they believe they could rely on their network partners to provide child care were more optimistic about being able to combine family and work.


In sum, these accounts suggest that the well-being of the respondents varied greatly. The continuum ranged from (a) those who reported that they had stable selfesteem, a harmonious family life, few conflicts with family and friends, the ability to manage well within their financial budget, confidence that they would be able to combine family and work, and optimism about finding a job in the near future, even though their current lack of employment was a source of stress; to (b) those who reported that they had low self-esteem, were burdened by multiple challenges, had a stressful family situation, were engaged in conflictual relationships, were struggling financially, and were resigned to the possibility that they would never be able to find an adequate job. Belonging to a vulnerable group is not automatically associated with having a low level of well-being; rather, it is a matter of how the individual balances her resources and her challenges. It may be assumed that well-being is related to education, duration of unemployment, and duration of separation. In these respects our respondents were very homogenous: they had similar educational levels (secondary education with or without a professional degree), and they had been unemployed and separated for more than 1 year. Interestingly, some of the respondents who had a professional degree reported experiencing severe strains and a low level of well-being, while some of the respondents who lacked a professional degree reported having a relatively high level of well-being. However, some of the biographical narrations reveal that the respondents found themselves in a vicious circle: most of those who left school early or broke off their job training had grown up in a very difficult family situation, had accumulated multiple strains over the years, and were currently in a difficult family situation. In terms of unemployment duration, the interviews indicate that a long duration of unemployment often coincided with multiple strains and a low level of well-being, with some of these challenges having been acquired independently of the employment situation, and others arising as a consequence of unemployment. The respondents also varied considerably in their perceptions of their separation from their partner: some described the separation as very painful and stressful, while others said it was the best decision they ever made. Some of the mothers found it difficult to specify a date of separation because it had been a long process of drifting apart or of separating and getting back together. Regardless of the amount of time that had passed since the separation, some of the respondents said they felt their situation had stabilized quickly, while others said they were still finding their way as a lone parent even several years after the separation.

In the following section, we will first look at the structural properties of the networks and present a typology, and will then focus on how the network types relate to well-being.

#### *Network Structure and Composition*

On average, the support networks consisted of six individuals. The smallest network contained only two supportive ties, while the largest network contained 16 supportive ties. The network density ranged from very sparse (0.17) to very dense, with all of the network partners knowing each other. On average, 2.73 of the supportive ties were multiplex, ranging from zero to six ties. The respondents reported providing support to zero to 11 persons (mean of 2.45). While some of these exchanges were reciprocal, many were not. The number of conflictual relations ranged from zero to four, with an average of one conflictual relation. This shows that the social networks of this specific group of unemployed lone mothers are very heterogeneous, displaying a broad variety of structural characteristics, while social isolation seems to be rare.

The support networks were composed mainly of relatives, friends, and institutional helpers. Eleven of the respondents included their current partner, while the children's fathers wer included only by four respondents. Other sources of support cited were acquaintances and the (ex-) partner's relatives (see Fig. 6.1). Eight respondents did not include any relatives, while six respondents did not include any friends.

The conflictual ties were mainly the respondents' own relatives (22% were sisters, 13% were brothers, 13% were mothers), their ex-partners (17%) and their

**Fig. 6.1** Composition of the support networks

ex-partners' relatives (13%). Friends, current partners, and neighbors were cited less frequently as conflictual ties.

#### *Network Typology*

We developed a network typology according to the networks' composition, looking at the type of ties that dominate the network. We found the respondents to be engaged in four different types of networks: (1) a network dominated by the family of origin, (2) a network centered on a conjugal relationship, (3) an extended network dominated by friends, or (4) a restricted network, which is very small and/or dominated by institutional helpers (see Table 6.1). We will describe the four types briefly in this section before giving more detailed examples of how they are related to wellbeing in the mixed-methods analysis in the following section.

In the networks dominated by the family of origin, relatives made up an average of 58% of the ties, with shares ranging from 40% to 80%. These networks were, on average, of medium size and high density, but their sizes and densities varied considerably (size: from four to 10; density: from 0.17 to 1). On average, 3.5 of the ties in the support networks were multiplex, ranging from one to four. These networks included 1.3 conflictual ties on average, and help was being provided to four ties on average.


**Table 6.1** Network types and their characteristics

For each measure the mean and (standard deviation) are indicated

In the second type of network, centered on a conjugal relationship, an average of 48% of the ties were to either the ex-partner or the new partner and his relatives, with shares ranging from 40% to 60%. In one case, both the ex-partner and the new partner were included. These networks were rather small on average, including just four or five supportive ties. Some were very dense (density of one), while others were extremely sparse (density of 0.17). The support network contained an average of three multiplex ties. The number of conflictual ties was comparatively high, and the number of ties to whom help was being provided was rather low.

The third type of network, the extended network, included a relatively high share of friends (on average 49%). These networks were very large, with a medium density. Many ties were multiplex, and only a few were conflictual. The number of individuals to whom help was being provided was very high. Two subgroups could be distinguished, (1) friendship-dominated networks of medium size, and (2) very large heterogeneous networks containing friends, relatives, and new or old partners and their relatives.

The fourth type is called a restricted network because these networks were small and sparse, and the support they provided was rather limited. Multiplexity was low, as was the number of ties to whom help was being provided. These networks had slightly more than the average number of conflictual ties. What is particularly interesting about these networks is that in most of them the share of network partners who were institutionalized helpers was very high (on average 41%). These helpers were mainly from the Youth Welfare Office or the Federal Employment Office, or they were social workers, family helpers, or mentors from job integration programs.

Figure 6.2 displays a typical example for the network structure of each type.

The network charts display the network partners, as well as their relationships to each other. A circle represents a woman, a triangle represents a man, while a square

**Fig. 6.2** Examples of the four network types

represents an institution or an institutional helper. A line indicates that the network partners know each other. All names have been anonymized and are fictious.

We are aware that a woman's experience of lone parenthood and unemployment is related to her education, her employment experiences, the duration of her lone parenthood, the ages of her children, and her relationship with the father of her

children. In our study, were these factors related to the type of network each woman was involved in? We would expect to find that women with higher levels of education, more job experience, older children, and a longer duration of lone parenthood would be more socially engaged because they had more opportunities to make contacts, more free time, and more time to adapt to lone parenthood and establish new contacts. When we look at the extended and the restricted network types, a clear picture emerges: extended networks were more common among respondents with upper secondary education and job experience, while restricted networks were more common among respondents with lower secondary education and long periods of unemployment. However, the ages of the children did not play the role we presumed: many of the respondents with pre-school children had an extended network. The characteristics of the respondents with the two other network types were rather mixed.

To sum up our findings on the respondents' networks, we can see that persons embedded in family-of-origin dominated and extended networks profit from a comparably large number of supporters, the contacts are often multiplex and the support is reciprocal. We therefore would assume that these network types coincide with high levels of well-being. Conjugal and restricted networks are rather small and therefore may not foster personal well-being as much. Additionally, the restricted networks often lack multiplex ties and include conflictual ties, which may result in a limited well-being. Our next analysis combines network and narrative data, in order to find out, if these assumptions hold and how certain network structures relate to personal well-being.

#### *How Is Well-Being Related to the Network Types and Characteristics?*

In our mixed-methods analysis, we confronted each respondents' network structure with her narration during the interview and analyzed how well-being and social relations are related. In this section, we present for each network type a) a case study that shows what resources the respondent had available, and how the respondent coped with critical events and chronic stresses and b) our general findings on the well-being of respondents embedded in this type of network. The case studies presented here correspond to the networks displayed in Fig. 6.2.

#### **Balancing Strains and Support in a Network Dominated by the Family of Origin**

#### **Susi Müller2 : "I am very content, because we help each other; it's a game of give and take"**

Susi Müller is a 33-year-old mother of three children (aged 10, 11, and 14). She has lower secondary education, was not trained in a job, and has never worked. Her husband was an alcoholic and the marriage was difficult. After the death of her husband 6 years ago, she moved back to the city her family lives in. She is currently involved in a new partnership. Her oldest child lives in a youth home. When she is asked why, she says that he refused to go to school and often got drunk.

Regarding all five areas we have identified earlier to pertain to well-being, she expresses content: Susi Müller is generally satisfied with her current life situation: "There are always ups and downs, but on the whole I am better than before (…) and since the death of my husband I have accomplished a lot." In terms of her family life, her employment status, and her finances she reports feeling content and sufficiently supported by her new partner, her mother, and her siblings: "If I have problems I can turn to my siblings; I have many siblings, I have nine siblings. And if something extreme happened, I would also turn to my mother. (…) She also takes the children to the doctor when I am working." Since the death of her husband, Susi Müller has stabilized her family life, has earned a school degree, and has participated in the federal volunteer service. She is currently getting her driver's license, and plans to do an internship in geriatric care. She thus sees herself as back on track, and looks optimistically to the future: "[In one year] I will be doing much much better."

2All names are anonymized.

The respondents embedded in networks dominated by the family of origin typically express high levels of well-being, also despite experiencing certain challenges (e.g., children in institutional care, financial problems, or diseases or disabilities of their own or of a close tie). Most of these respondents, such as Susi Müller, reported that they were handling these strains well with the support of their family, and expressed a relatively high level of life satisfaction. A few said the support available from their family was not sufficient, either because their family members were too poor to provide adequate financial aid, or were too sick to provide adequate instrumental support. These respondents said they perceived their everyday life as stressful, and found combining family and work difficult or even impossible. From this we infer that being embedded in a network with family members does not automatically coincide with well-being, but fosters well-being only if these family members are able to provide adequate support. The most supportive person among the relatives is typically the mother, but siblings could also take this position. Interestingly, some of the respondents did not really return to their family of origin after separating from their partner, as they had never truly left home. For example, in some cases the woman had her first child at a young age, separated from her partner soon after conception, and had relied on her family for help with raising her oldest child ever since.

#### **Balancing Strains and Support in a Conjugal Network**

#### **Sabrina Meier: "My network is sufficient"**

Sabrina Meier is a 28-year-old mother of three children (aged five, seven, and nine). She holds a lower secondary degree, has not finished her job training, and has no work experience. This month she starts working in a mini-job. She separated from the children's father two years ago, and found the separation extremely difficult and stressful. She moved with her children to another city and argued with the children's father about custody. During this period, the younger children were living with foster parents. One year ago, the situation was stabilized and settled by the court: the youngest child lives with the father and the two older girls live with her. She has also had a new boyfriend for the past year and a half.

In general, she feels satisfied with her life, and believes that she has finally managed the separation from her partner. She is content with her family life and looks forward to moving in with her boyfriend. She feels supported by her boyfriend, his mother and sister, and her own sister: "My sister, his mother, his sister or his mother's partner [support me], depending on who is available. There is always somebody who can help, because we all help each other." She perceives contact with the children's father as conflictual. Being unemployed is stressful for her, as she has nothing to do when the children are in school. Now, however, she has a dog. She is pleased that she found a mini-job, and hopes that she will soon have more stable employment: "Things are looking up."

Like all of the respondents who were embedded in conjugal networks, Sabrina reported having a high level of well-being, despite also having various problems. Although these respondents said they perceived their joblessness as stressful, they also reported that they were largely content with their family life, and that the support they received from their partner and their partner's family offset their stresses and strains. Only in case the conjugal family members were unable to provide adequate support e.g. due to chronic/severe diseases, these networks did not foster wellbeing as much. Persons embedded in this type of network have reduced or broken contact to their family of origin, mostly because the relationships had been very conflictual in the past. The respondents have compensated this loss by engaging into contacts with their partner's family. The conjugal networks show that the partner's family can be as supportive as the family of origin. Regarding the size, a medium size seems to be enough to foster well-being, and characteristics as multiplexity seem to be more important than the size. However, as we noted above, the mere existence of supportive conjugal family members is not necessarily an indicator of the availability of support adequate and sufficient in order to foster well-being.

#### **High Level of Well-Being in an Extended Network**

#### **Doreen Schulz: "On the whole I am content. The only thing making me discontent is searching for a job"**

Doreen Schulz is a 35-year-old mother of a five-year-old child. She holds an upper secondary vocational degree and has work experience, but not in the job she was trained in. She has been unemployed for 2 years. She did not experience the separation from the child's father as a sudden or significant break, and therefore cannot really give a date of separation. Instead, she says, the process of separation started with the arrival of the child (which was planned), as the child's father has never really built a relationship to his child nor participated in family life. She was therefore used to organizing her life on her own for a long time before he finally moved out. Today she has a good relationship with her child's father.

She describes herself as content with her general life situation. In terms of the life domains, she expresses positive feelings and a high level of satisfaction with her family life and does not report any strains associated with being a single mother. She feels supported by her family (her parents and her cousin help her take care of the child), and reports that her mother has taken over much of the function of the child's father: "For my daughter the second attachment figure is not daddy, but grandma." She can fully rely on her mother's support, and they are in close and frequent contact. For example, her daughter calls her grandparents every evening to wish them a good night. She has many friends, some of whom are old school mates, and some of whom are mothers of her child's friends. She is not satisfied with her lack of a job: "It's a little frustrating, that one keeps applying for jobs and only gets letters of refusal or no answer at all." She is frustrated about being home during the day while her friends are all working. In this situation as well she feels supported by her family and friends, who tell her about open positions. Her mother has offered to help her with child care if she returns to full-time work. She is optimistic about her prospects of finding a new job, and expects to be able to combine family and work with the help of her mother. In sum, she does not perceive lone motherhood as stressful, and she does not report any major strains or stresses apart from being unemployed. She describes her social network as supportive, providing emotional, instrumental, and informational support, as well as social companionship. She does not have any conflictual ties.

Doreen Schulz was typical for all cases in this category: they had upper secondary vocational degrees, job experience, a relatively high level of general life satisfaction, relatively few strains and stresses, and a variety of supportive relations they could rely on. These women either never had or were largely able to free themselves from conflictual ties by ending their relationship, which positively affected their well-being. They reported that friends were helping them manage family strains by providing emotional closeness, sociability, and various forms of support, but offered only limited support in terms of child care. In case their extended networks were made up entirely of friends and did not include any family members, these networks did not provide adequate child care support –as this kind of support is most helpful when it is provided regularly, flexibly, and without asking for too much help in return. Although the respondents who were embedded in friendship-dominated networks were generally optimistic about finding a new job, they also said they perceived combining family and work as very difficult and stressful. From this we infer that family members play a crucial role in child care support, and cannot easily be replaced by friends.

#### **Limited Well-Being in a Restricted Network**

#### **Mareike Lehr: "Friends – I don't really have any. (…) About personal matters I talk to pals from the internet"**

Mareike Lehr is a 26-year-old mother of a six-year-old child. She holds an upper secondary vocational degree, and has been unemployed since she finished her job training. She separated from her child's father 5 years ago, and has had three relationships since then. Two years ago, after a relationship of 18 months ended, she fell into a depression. She is currently in the process of starting a new relationship.

She expresses very limited life satisfaction. Her biographical narrations reveal various strains: a difficult relationship with her mother and grandmother, violence in a former partnership, depression, difficulties in caring adequately for her child, and debts. She loves her child, but she is also happy when she can leave him with her grandmother, even though the relationship with her grandmother is also sometimes conflictual. Regarding family life, she is grateful for the support of two institutional helpers. Although she is aware that it is their job to help her, she states that she has "the feeling, this is like a good friend." Asked about friends, she states that she does not have any, only some "pals", that is mainly contacts via the internet. She does not perceive her lack of employment as stressful, but her financial situation leaves her with little means to pursue free time activities. She has few expectations for the future; her planning horizon extends only over the next few weeks and months. For example, because the period during which she receives institutional help is running out she is thinking about letting her household become a mess again so that the assistance is extended. Her network contains her grandmother, her mother, and her new boyfriend, as well as four institutional supporters who help her with her family and with finding a job.

Most of the respondents with restricted networks reported having severe and multiple strains. They indicated that they had low levels of well-being in terms of their everyday family life and their employment situation. They reported having low self-esteem, financial problems, and/or conflictual relationships. When asked about their networks, they cited few private contacts, and often could not name any close friends. The strains they were experiencing appear to have hampered them in engaging in social relationships. For example, having to spend large amounts of time caring for a disabled child or being severely depressed restricted their ability to engage in reciprocal relationships.

The existing network partners of these women were not able to provide adequate support, either because the amount of support needed was extremely large, or because the supporters had little time, money, or strength themselves; their positive impact on well-being was therefore very limited. In some of these networks, institutional helpers stepped in and took over supportive functions. While some of these networks were of women in a state of transition (they had just moved and had dissociated themselves from their former network partners, but had not yet been able to build new contacts), others were networks of women who were under extreme pressure for a long time, and who had little access to social resources.

#### **Discussion**

Research on lone mothers has long taken a deficit perspective, mainly looking at what single mothers are missing, and describing them as facing a high risk of social isolation. However, newer research has stressed the agency of lone mothers in coping with their life situations, and reveals a more differentiated picture. These studies have applied a network perspective and have thus looked in detail at the personal relations lone mothers are engaged in, as well as at the structural characteristics of the networks they form. They have found that lone mothers have a high level of social integration and access to support. However, many of these studies focused on employed women. What would we find if we looked at unemployed lone mothers? Is the hypothesis that single mothers are socially isolated more applicable to this group because they are neither integrated into a family-in-law nor into employment, have to deal with raising their children and finding a job on their own, and are at high risk of poverty? Moreover, would the stressful life situations of these lone mothers lead them to have looser contacts, or to have relationships with people who are in a similarly strained situation, and are therefore unable to provide much support? Alternatively, would these lone mothers need more support than their network partners could provide?

Interestingly, our findings indicate that the hypothesis of social isolation does not hold for this group either, and that the picture is much more differentiated. The social networks of our respondents ranged from being very small to being very large, and the amount of support they reported receiving ranged from being very little to a lot, and from insufficient to adequate. All of the respondents said they received some form of support.

To better capture the variety of networks available we developed a typology in which we distinguished between four network types according to the network composition: networks dominated by the family of origin, networks centered on a conjugal relationship, extended networks, and restricted networks. The first type, or networks dominated by the family of origin, were similar to the "return to family of origin" type described by McLanahan, Wedemeer, and Adelberg (1981) and the "family network" type presented by Niepel (1994). Our type was, however, structurally different from these types because the networks' sizes and densities varied considerably, and were not all very small and very dense. While McLanahan, Wedemeer, and Adelberg (1981) described their respondents as being in a "family haven," and having little contact to peers or engaging in social activities outside the family, most of our respondents had friends, and their networks were not necessarily small or dense. However, unlike in the extended networks, in the networks dominated by the family of origin family members provided most of the support.

The networks centered on conjugal relationships (dominated by the (ex-) partner's relatives) were similar in composition to the corresponding networks described by McLanahan, Wedemeer, and Adelberg (1981). The structure was somewhat similar, as many of these networks were rather small, sparse, and dense.

The extended networks we described were similar in composition and structure (large in size, and composed of various groups, such as relatives, friends, and the partner's relatives) to the extended networks found by McLanahan, Wedemeer, and Adelberg (1981) and the family and friendship networks identified by Niepel (1994).

Our restricted networks did not match any of the network types identified by McLanahan et al. (1981) or by Niepel (1994). This network type was found mainly among mothers who were experiencing severe strains or multiple problems, and were therefore highly vulnerable. These networks were mostly small and sparse, like the friendship networks described by Niepel; but they included relatives and hardly any friends. The women who had these networks were socially isolated in terms of their private contacts, but institutional helpers who served as emotionally and instrumentally supportive ties sometimes filled this gap. The fact that these respondents had a large number of institutionalized helpers could be due to our sampling strategy, as we recruited women who were enrolled in a program for labor market re-integration. However, many of the institutional helpers mentioned were employees not only of these programs, but of institutions dedicated to helping people stabilize their family life. This institutional support can play a pivotal role for women embedded in restricted networks, as it can provide these women with a minimum level of support, help them stabilize their living situation, and get their life back under control after a series of setbacks.

Returning to our question of whether lone mothers are lonely, we can state that even within the vulnerable group of unemployed lone mothers, all of the women were somewhat socially integrated and received some form of support; thus, they were not fully isolated, but those embedded in restricted networks had a very limited social safety net. Although the women who were embedded in the other three

Our respondents, who were recruited from an employment program, were willing and able to participate, to accept help, and to be interviewed. It is likely that there are more socially isolated lone mothers who cannot or do not want to attend to such a program, and who would not consent to being interviewed. We know little about the support these highly vulnerable people receive, or about their coping strategies.

Regarding the impact of the network types on well-being, McLanahan et al. (1981) stressed that the relationship is quite complex: "Each network type appeared to provide certain types of support which were adaptive for certain groups of single mothers." They also noted that the effects of social structure and social support on well-being are moderated by a third factor: the mother's role orientation. They identified "stabilizers" who wanted to maintain their pre-divorce roles (wife or mother), and "changers" who attempted to establish a new identity (generally a role connected to their employment career). Stabilizers and changers tended to build different types of networks, and appeared to benefit from different network types. The researchers argued that it is network density that matters most: sparse networks improve well-being for women who are "changers," while dense networks improve well-being for "stabilizers."

In our study, which was conducted more than 30 years later and on respondents from eastern Germany, these role relations did not apply. To most eastern German mothers combining work with family is a given. We thus found that the inclination to work was extremely high: even among our sample of long-term unemployed mothers who had often failed in finding or keeping a job, only a few had resigned themselves to being unemployed.

Our findings suggest that different network types were associated with differences in well-being. At the top of the scale were extended networks, which were highly supportive, and were found among women who said they felt content with their family life and were optimistic that they would soon find a job. At the lowest end of the scale were restricted networks, which did not offer adequate support and were found among women who said they felt stressed in their family life, and that they had resigned themselves to being unable to find a job.

The association between network structure and well-being was bidirectional: (1) the network type was a consequence of the strains the respondents were experiencing, and (2) the extent to which social networks were able to foster well-being depended on their structure and composition, the resources the network partners were able and willing to supply and how the respondents evaluated the support.

If we look at this bidirectional influence by comparing the two types of networks that were the most different—i.e., extended and restricted—we can see that the women who had low levels of well-being and were experiencing severe strains were not able to invest in their network, to form new relationships, or to provide help in return for the support they received. Thus, their contacts mostly provided one type of support, the multiplexity of their contacts was low, and their networks were restricted. The support provided in these restricted networks was often insufficient, and the social relations did little to improve the women's well-being. Many of these women had biographical strains, such as a difficult childhood, that had long-term negative effects on their well-being as an adult, and led them to cut ties with their own relatives. In contrast, respondents who were embedded in extended networks reported having a relatively high level of well-being and a greater willingness and ability to make new contacts and to invest in their relationships, by, for example, providing help to their network partners. Their many network partners provided them with various forms of support in return. Because their ties were often multiplex, they could generally rely on a range of network partners to provide them with support, which helped to relieve their stress and fostered a high level of well-being. This appears to be an example of the Matthew effect: having a high level of wellbeing leads a person to form supportive networks, which in turn fosters the person's well-being. This effect can also work in the other direction: having a low level of well-being leads a person to have a limited network of partners who cannot provide the person with adequate support to foster her well-being. This can become a vicious circle that is unlikely to be broken without the intervention of a third party, e.g. social workers.

From a structural point of view, the question of whether supportive networks have certain structural features arises. In terms of composition, we found in our sample that extended networks that were of medium or large size and of medium density, and that contained heterogeneous network members (mainly the woman's own relatives, her partner's relatives, and her friends), a large number of multiplex ties, and many reciprocal relations were the most supportive. We also found that family-dominated networks and networks centered on conjugal relationships (of small or medium size, of high density, with medium multiplexity and low or high levels of reciprocity) provided adequate support, but that they were more vulnerable: if the woman experienced severe strains or the relative or partner was unable to provide support (e.g., in case of sickness), she often had few or no network partners who could assist or replace the supportive tie. Interestingly, we found that extended networks without any relatives who are supportive in childcare could not adequately help to combine family and job, which hampered well-being. This stresses the important and beneficial role of own or the partner's relatives for regular and reliable support in childcare. Least supportive were the restricted networks, which were of small size, and multiplexity, and had few reciprocal ties. In our sample, we found that institutional helpers sometimes stepped in to shield respondents from social isolation, but that these helpers could not replace the emotional support and stability provided by personal private ties.

To sum up, in terms of network structure, it is not density alone, as proposed by McLanahan, Wedemeer, and Adelberg (1981), that determines the supportive effect of unemployed lone mother's networks. In our research, we found that the most supportive networks were of a large or a medium size, and included heterogeneous and multiplex ties (including the woman's own or her partner's relatives). Trying to identify the structural properties of the different network types does not provide a clear picture. While our findings are similar to those of previous studies in some respects, they also differ considerably from them in others. To learn more about typical networks and the correspondence between network types and their structural properties, future research should include additional network measures and involve a larger sample in order to explore in more detail how structural features and wellbeing are related.

Our research shows that beyond the networks' structural properties, also their meaning to the respondents had to be taken into account: only if the ties are not perceived as conflictual and if the support provided is evaluated as adequate, these ties can foster well-being. As the associations between structure, support, and wellbeing are rather complex, future research on these links is needed, which includes not only structural information but also the respondents' subjective evaluation of their ties and the support available.

Although we had retrospective data, our research was limited by its crosssectional design. To better disentangle the causal relationship between social networks and well-being, we would need longitudinal data that cover changes in the partnership status as well as in the social network structure. Our retrospective narrations captured some of the changes, but could not fully represent the dynamics in network structures over time.

Another limitation of our research is that we focused on a group of lone mothers who were long-term unemployed, even though in Mecklenburg-Western Pomerania the majority (65%) of lone mothers are engaged in paid work (Buch et al. 2012: 20). We chose to focus on this group because we assumed that doing so would provide us with the best chances of detecting vulnerability; an issue we wanted to analyze in depth. Future research should extend the analyses to the social networks of employed lone mothers, as well as (employed and unemployed) partnered mothers. Moreover, given the persisting differences between eastern and western Germany in terms of both the labor market and family relations, the view should be broadened to western Germany. As the existing literature on lone fathers is very limited, more research on their social networks and coping strategies should also be conducted in order to learn more about gender-specific differences in social networks and well-being.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **References**


**Dr. Sylvia Keim** is a sociologist and research associate at the Institute of Sociology and Demography, University of Rostock, Germany. Her research interests include sociology of the family and the life course, social networks, social and health inequalities, qualitative research methods and mixed-methods research designs. Her relevant publications include: (with L. Bernardi and A. Klärner) Social Networks, Social Influence, and Fertility in Germany: Challenges and Benefits of Applying a Parallel Mixed Methods Design, in: Dominguez, S.; Hollstein, B. (eds.), Mixed Methods Social Networks Research. Design and Applications. Cambridge University Press (2014); (with L. Bernardi): In the early thirties and still without children? Life paths and family models of working women in East and West Germany, in: Kreyenfeld, M.; Konietzka, D. (eds.), Childlessness in Europe. Patterns, Contexts, Causes, and Consequences. Springer Verlag (2017).

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Chapter 7 Migrant Status and Lone Motherhood – Risk Factors of Female Labour Force Participation in Switzerland**

**Nadja Milewski, Emanuela Struffolino, and Laura Bernardi**

#### **Introduction**

The focus of this chapter is labour market participation among lone immigrant mothers in Switzerland. Lone immigrant mothers are exposed to four stratification risks on the labour market. First, women are generally disadvantaged relative to men. Second, mothers of young children are less active on the labour market than women in general (Boeckmann et al. 2015). Third, lone mothers who work may experience particularly high levels of role incompatibility as workers and primary care-givers. Fourth, international migrants have fewer labour market opportunities than non-migrants, and this disadvantage continues into the second generation (e.g., Constant and Massey 2005).

Despite the fact that they are exposed to multiple risk factors, lone parenthood among international migrants has not received much attention in the literature. Little is known about the patterns, the determinants, and the consequences of lone motherhood among migrant populations, including about their labour market activity. There are several possible reasons for this research gap, such as that the integration literature has focused mainly on male migrants, as men were assumed to be the initial migrants and the economically (more) important members of migrant families (Lutz 2007; Pries 2010; Charsley 2012). Studies in family demography have mainly focused on the processes of family formation among immigrants living in

N. Milewski (\*)

University of Rostock, Rostock, Germany e-mail: nadja.milewski@uni-rostock.de

E. Struffolino WZB Berlin, Humboldt University, Berlin, Germany

L. Bernardi

National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life course perspectives (NCCR Lives), University of Lausanne, Lausanne, Switzerland

western Europe: i.e., union formation, marriage, transition to the first birth, and subsequent childbearing.

Only in recent years has union dissolution among migrants become a topic in demographic studies on immigrant groups living in western European countries (Kulu and Gonzáles-Ferrer 2014). The literature on family formation suggests that lone parenthood has been rather rare among immigrant groups, because, for example, international migration policies tend to favour married family migrants. Therefore, the share of people who are married has long been higher among migrant groups than among the non-migrant population in the respective destination country. Union stability among immigrants has also been rather high due to their high marriage rates and relatively large numbers of children. Rates of extra-marital childbearing and of divorce have been low in immigrant populations (e.g., Milewski and Kulu 2014).

In recent years, however, the nature of immigration has changed. The percentage of women migrating – whether alone, as the breadwinner, or as the forerunner in the family – has increased (Lutz 2007). At the same time, the percentage of immigrant descendants in the total population has been rising in many western European countries. These processes of population change appear to be accompanied by tendencies among migrants to adapt to the demographic behavioural patterns of the host country: the trends towards non-marital cohabitation, extra-marital childbearing, and marriage/union dissolution that were observed in non-migrant populations some two decades earlier have recently been spreading to migrant groups (Staat and Wagenhals 1996). Consequently, lone parenthood is becoming increasingly common in migrant populations.

We examine the labour market activity patterns among immigrant mothers while taking into account their partnership status. Over time, the literature has showed consistently and with only a small amount of variation by country that lone mothers are, on average, at higher risk than partnered mothers of being poor, of being dependent on welfare, of being unemployed, of having low earnings, and of having high child care costs (Staat and Wagenhals 1996; Zagel 2014). We add to the literature by focusing on the intersection of partnership and migration status, and specifically by investigating whether being a lone mother and a migrant at the same time represents an additive risk for non-employment. We also account for within-group variation by differentiating between the immigrant generations. In addition, we explore the heterogeneity across migrants by their countries of origin.

We carry out our study on immigrant mothers living in Switzerland, a multilingual country with a long immigration history. On average, migrant women are more likely than non-migrant women to be in full-time employment when they work, and to be dependent on social welfare (Liebig et al. 2012). This is also the case in Switzerland, a country with an extremely diversified foreign-born population in terms of geographical origin, socio-economic position, and migration trajectory (Afonso 2004; Fibbi and Wanner 2009; Laganà et al. 2014). At the time of writing, the largest immigrant group in Switzerland is from Italy, followed by groups from Germany, Portugal, and the former Yugoslavia (FSO 2015). The last two groups are the most recent arrivals. The Portuguese migrants are mainly economic migrants from within the Schengen area of free labour market circulation. The migrants from the former Yugoslavia are rather heterogeneous: some are economic migrants, others are refugees who were fleeing forced army recruitment or ethnic persecution (especially from Kosovo), and more recent migrants came through family reunification. Today, the percentage of the country's population who are foreigners (i.e., who are not Swiss citizens) is about 22%. About 27% of the total population of Switzerland are foreign-born; one of the highest shares in Europe (OECD 2015).

Overall, female labour force participation (FLFP) levels are rather high in Switzerland: about 81% of the native-born women and about 76% of the foreignborn women are employed (OECD 2015). However, because of the low availability of public child care and the resulting family-work conflicts, a large portion of women in Switzerland work part time (Bühlmann et al. 2012). Weak labour market attachment, poverty, and welfare dependency are major concerns in the sociopolitical debate about lone mothers (Salter 2017), as well as about migrants. Looking at the labour market situations of immigrant lone mothers can thus provide insights into the often precarious living conditions of mothers who are primary care-givers and income providers, and who may also face specific difficulties related to their immigrant status.

#### **Background**

#### *The Swiss Context and Lone Mothers' Employment*

According to official Swiss statistics for 2012, 15% of the households with at least one child under age 25 were one-parent households (FSO 2013). Individual data from the *Enquête sur les familles et les générations* (EFG) 2013 indicate that about 6% of the total population between ages 15 and 55 were lone mothers living with at least one biological child below the age of 18 (Struffolino and Bernardi 2016). While these numbers present a snapshot at the moment of the survey, retrospective data on the union and fertility histories of the women in the sample indicate that the prevalence of lone parenthood is higher: almost 13% of these respondents had experienced lone motherhood between 1953 and 2013. Some of these women "exited" lone parenthood because their children's biological father or a new partner moved into the household, while for others lone motherhood ended because their grown-up child(ren) left the household. A one-parent household has become a relatively common living arrangement that growing numbers of individuals are experiencing for durations ranging from a few months to several years (Struffolino and Bernardi 2016). The great majority of lone parents are women (89% in 2000, cf. Bühler 2002) who become the primary care-giver and the main breadwinner of their family.

In Switzerland, the average income of lone parent households is far lower than that of other household types, with the exception of households of people aged 65+ who are living alone (OECD 2015). The high poverty risk faced by lone-parent households can be partially explained by the gendered structure of the labour market (Flückiger 1998; Branger 2003) and the disadvantages women experience when they are active in the labour market, including work-family conflicts and a tendency to work part time (FSO 2013).

On the one hand, Switzerland offers relatively widespread social policy protections against poverty (Bertozzi et al. 2005; Armingeon et al. 2004) and rather generous unemployment benefits (Gebel 2013). On the other hand, the care system is mostly relegated to private institutions, and the public policies targeting work-family reconciliation are less robust than those of other western welfare states (Monnier 2006): school schedules are hardly compatible with those of parents who work, and expensive public child care and marriage-based taxation discourage women's labour force participation (Bütler and Ruesch 2007). These welfare arrangements are based on the one-and-a-half-earner model, in which men work full time while women who become mothers adjust their working hours to accommodate their care obligations (Giraud and Lucas 2009). As a consequence, while Switzerland has a higher female labour market participation rate than many other European countries, most Swiss women are employed part time (Bühlmann et al. 2012).

"Despite women's inroads into employment and men's increasing participation in childcare, women remain primarily responsible for children. Thus, motherhood is an axis of inequality central to our understanding of the processes that shape women's employment patterns" (Boeckmann et al. 2015: 2). Lone mothers are likely to be particularly disadvantaged, because there is no partner in the household to compensate for their relatively low income (Stutz and Knupfer 2012). Therefore, the major challenge for women in Switzerland, where the care and work spheres are strongly gendered, is to increase their disposable income after making the transition to lone parenthood, while continuing to devote time to care. However, as social protections may cover the most immediate and urgent economic needs of lone mothers, those with unfavourable labour market prospects may have a lower incentive to take up paid work (Gangl 2006; Luijkx and Wolbers 2009; Dieckhoff 2011).

Hence, our first working hypothesis concerns the effect of the partnership status on labour market participation: *Lone mothers are more likely to either work full time or to be out of employment than mothers who share their household with a partner (H1).*

#### *The Role of the Migrant Status for Labour Market Participation*

The literature on the labour market activities of women of majority populations has mainly focused on women's efforts to reconcile work and family life, the wage gap between men and women, and the wage gap between women based on their motherhood status (e.g., Boeckmann et al. 2015). Although these studies took into account cultural factors by comparing different countries, most did not account for cultural heterogeneity within a single national population by controlling for immigrant status or citizenship.

In research on international migrants, structural integration is one of the most prominent research areas, with education and labour market being the most important and probably most frequently studied domains. The literature on labour market participation among international migrants has generally assumed that the crucial disadvantage faced by women is the high risk that they will be excluded from the labour market entirely; whereas the main disadvantages faced by men are the high risk of unemployment, of having a low income, and of having a low-level occupation due to various mechanisms of selection into migration and of integration (Algan et al. 2010; Höhne and Koopmans 2010).

For first-generation migrants' labour market integration of, human capital understood as the set of worker's marketable skills in which workers make a variety of investments (Becker 1993)—plays a central role. According to the labour market assimilation theory (Chiswick 1978), immigrants may have initial disadvantages relative to non-migrants, though these disadvantages may vary by country of origin and destination; and these disadvantages may decrease or level off with increasing length of stay. The disadvantages are mainly related to the devaluation of foreign educational degrees, the relatively low host-country specific knowledge (e.g., the language proficiency), the negative selection into migration, and the preferences of migrants and their motivations for return migration or intentions to stay; i.e., the intention to return is negatively associated with an investment in host countryspecific knowledge and skills (Kalter 2008; van Tubergen 2008).

For migrant women in western and southern European destination countries, the empirical evidence on their labour force participation tends to confirm the labour market assimilation theory. The migrants' FLFP rates increase with length of stay, and eventually reach about the same rates as those of the respective majority population (Rendall et al. 2010). The empirical evidence on the economic success of migrant women cannot fully be explained by this theory, however. Increasing attention has been paid to the role of the partner; i.e., the influence the husband's human capital has on his wife's labour market career (Lin 1999; van Tubergen 2008). The results of these studies were in line with new home economics and the household specialisation theory (Becker 1991), which states that household's members specialize in the production of commodities to maximize the household utility function, that is through the division of labour between the household and the labour market. Long (1980), for example, showed that the income of immigrant women in the US decreases with increasing length of stay: a woman may participate in the labour market immediately after migration, but gradually withdraw when her partner has gained host country-specific human capital and has improved his own status on the labour market. Duleep (1998) called this the "family investment hypothesis".

FLFP is an indicator for the integration of immigrant women and for cultural transformation processes, since the labour market activity of women is influenced by cultural and/or religious norms concerning gender roles, traditional patterns in FLFP in the respective country of origin, as well as the size and the structure of the family (Reimers 1985; Antecol 2000; Blau et al. 2013). Previous research on international migrant women has mainly been concerned with gender roles and the division of labour in the family, and the differences between the gendered employment patterns in the women's respective counties of origin and of destination. More recent work has taken into account within-group variation by differentiating between the immigrant generations and/or the respective countries of origin (e.g., Blau et al. 2013; Milewski 2013). Studies that further differentiated the employment status or the type of work among immigrant women tended to focus on precarious job conditions, such as the informal sector in domestic services or prostitution (Han 2003).

The literature based on representative surveys paid little attention to immigrant lone mothers until very recently, given their relatively low numbers in the samples and the relative difficulties to perform detailed analyses on them (we will come back to this in section 3 of our paper, where we describe the data).

In Switzerland, the labour market participation patterns of immigrant women differ substantially from those of Swiss women: On the one hand, immigrant women have a much higher unemployment rate than Swiss women (9% versus 2%), though this difference may be partially due to inactivity among Swiss women, who may not search for a job if they are less attached to the labour market than migrant women (Winkelmann 2002). On the other hand, there are differences when we take into account the employment patterns: compared to their native-born peers, immigrant women are equally likely to be in full-time employment and less likely to be in parttime employment. For example, in 2009 among all women of working ages, the shares of women who were in full-time employment were 31% for native-born women and 32% for foreign-born women from lower-income countries; whereas the shares of women in part-time employment were 44% and 27%, respectively (Liebig et al. 2012). Thus, the average difference in employment rates between native-born and foreign-born women from lower-income countries has been traced back to the higher rates of part-time employment among non-migrants (Liebig et al. 2012).

Nevertheless, evidence on skilled migrant women from Latin America, Asia, and south-eastern Europe from studies using qualitative methods has shown that most of these women could not use their social and cultural capital to establish themselves in the stronger segments of the labour market, and that a considerable share of them are either excluded from the labour market or employed in jobs for which they are overqualified. Only one-third of these women report working at their skill level, and then only irregularly (Riaño and Baghdadi 2007). The evidence on the wage differentials between immigrants and non-migrants is contradictory, which suggests that the size of the gap varies depending on the migrant group and their composition by education. These differentials are substantially smaller among second-generation migrants (de Coulon 2003). A recent investigation on the positions of immigrants in the labour market in Switzerland concluded that the size of the gap between migrants and Swiss natives mostly depends on the migrant group and on whether the migrants completed their education before or after arriving in Switzerland (first-generation migrants). Differences in the labour market positions of natives and migrants tend to fade among second-generation migrants (Guarin and Rousseaux forthcoming). Other studies emphasize variation by country of origin though (Laganà et al. 2014).

Based on the theoretical considerations and the previous evidence on the labour market participation patterns of immigrant women, we have formulated hypotheses concerning their participation/non-employment levels and types of labour market activity by taking into account their employment hours: *(H2) following the human capital approach, we assume that non-employment rates are higher among migrant lone mothers than among non-migrants due to the general disadvantages of migrants in the labour market. Among those who are gainfully employed, we expect to find* 

*that migrants on average work longer hours than non-migrants due to their lower wage returns.*

Moreover, we expect to find variation between migrant groups; i.e., that the employment patterns vary between first-generation migrants and their descendants, as well as between migrants based on their country of origin: *(H3) in line with the human capital approach, we assume that the FLFP patterns of second-generation migrants are more similar to those of non-migrants than to those of the first generation, since their host country-specific education and knowledge levels are higher than those of the first generation* (Becker 1991; van Tubergen 2008).

#### *Socio-demographic Factors*

Finally, we consider what role individual and household characteristics play in shaping the differences between lone mothers and mothers in couples, as well as between migrants and non-migrants: if there are any differences in the employment patterns of the respective groups, *we expect to find that they are smaller when we control for the composition of these groups; i.e., for the socio-economic characteristics of the mother and the household characteristics (H4)* (Laganà et al. 2014).

#### **Data and Methods**

#### *Data*

For our analyses we used data from the Swiss Labour Force Survey (SLFS). This cross-sectional survey collects information about labour market participation, and offers a sample size that is large enough to allow us to compare lone mothers and mothers in couples by migrant status. We used the 2008 wave, as this was the last wave in which the questions concerning the country of birth of the parents of the respondents were included. Thus, for the respondents of the survey waves from 2009 onwards, we would have been unable to determine to which migrant generation they belonged. We selected all of the women who were between 20 and 54 years old and who were living with at least one child under age 18. We defined as lone mothers those mothers who were not sharing their household with a partner. Mothers in couples were defined as those who were sharing their household with a partner, irrespective of their marital status (and no matter whether the partner was the biological father of the children in the household).

The respondents in our sample were defined as non-migrants if they and both of their parents had been born in Switzerland, and as migrants (who were oversampled in the data) if they or either of their parents had been born outside of Switzerland. The final sample consisted of 4617 migrants and of 2197 non-migrants. In order to take into account within-group variation among migrants, we first distinguished between the migrant generations: individuals were defined as belonging to the first

migrant generation if they were living in Switzerland in 2008 but had been born in another country to parents born abroad, and had migrated at age 15 or older. Secondgeneration migrants were defined as individuals who were either born in Switzerland to one or two parents born abroad, or who were born in another country to parents born abroad, and who had migrated before age 15. (The age used to distinguish between the migrant generations varies in the literature by research topic. In addition, the size of the respective sub-samples matters for statistical analyses. Since in Europe second-generation migrants tend to be rather young, we chose to group together those migrants who moved as children and those who were born in Switzerland. This seems appropriate for our research question since they all participated in public schooling in Switzerland, and lived there during childhood and/or adolescence. Hence, the second-generation migrants all experienced the Swiss societal context before the start of their reproductive years and of a labour market activity. Because the first-generation migrants had lived exclusively in their country of origin before age 15, their family formation behaviour may be more influenced by the context of their country of origin (Rumbaut 2004; de Valk and Milewski 2011).) About 70% of the migrants in our sample belonged to the first generation, while the remainder belonged to the second generation.

Second, we identified migrants according to their country of origin (without distinguishing between the migrant generations due to very small sample size for lone migrant mothers). The countries of origin were grouped into five categories: Switzerland (i.e., non-migrants), Western Europe (including Nordic countries), Eastern Europe, Southern Europe, and all other countries. If they were born in Switzerland, the country of origin of the second-generation migrants corresponds to their parents' country of origin. Such grouping reflects, as much as it was possible given the small sample size, the patterns of FLFP across regions, which are lowest in Mediterranean countries and highest Eastern Europe (Boeckmann et al. 2015).

Of the women in our sample, 10.5% were lone mothers and 89.5% were mothers in couples. These numbers are consistent with the census data from the Swiss statistical office on the family composition of households with at least one child (FSO 2015).

In our sample, 11% of the non-migrants and 10% of the migrants were lone mothers. Since there is no literature on this topic for migrants in Switzerland, we cannot compare these numbers with figures from other publications. We calculated, however, the respective numbers using two other longitudinal and representative datasets that contain immigrants living in Switzerland and non-migrants, and that are often used for family-related research questions: the Swiss Household Panel (SHP) and the *Enquête sur les familles et les générations,* a Swiss survey based on the Gender and Generation Survey (GGS) questionnaire design. Although these surveys provide a wide range of information on individuals' lives that we could not find in the SLFS data, their sample sizes for specific population groups are small. For both data sets we estimated consistent shares of lone parent families, but our samples were extremely small when we distinguished between non-migrants and migrants, and were even smaller when we tried to explore different migrant categories in greater depth. Given the lack of any kind of empirical evidence on lone parenthood and migrant background, we chose to use the SLFS, which provided us with a reasonably large sample size to empirically test our hypotheses. Yet, since the SLFS retrospectively collects employment and family data, the prevalence of ever-experiencing lone motherhood may be slightly higher (Struffolino and Bernardi 2016).

#### *Dependent Variables and Method*

We first ran a set of binomial logit regression models estimating the probability of being employed versus being out of employment1 . We then carried out multinomial logit regression models in order to differentiate between the labour market participation levels; we estimated the probability of being unemployed/inactive, in fulltime employment (>90% of a 40-hour working week), in long part-time employment (50–89%), or in short part-time employment (<50%). For both analyses, the independent variables of interest are the interaction between partnership status (mothers in couples or lone mothers) and a type of migrant background, operationalised as either the immigrant generation or the country of origin.

The results are displayed as predicted probabilities for the independent variables, and are to be interpreted as the linear probability connected to the respective category or the interaction of interest for all possible outcomes, net of the variables the models control for (Wooldridge 2002; Long and Freese 2014)2 .

#### *Control Variables and Sample Description*

Following the literature on labour market participation among (lone) mothers (Baker et al. 1999; Boeckmann et al. 2015), we included a number of control variables and possible confounders of the main association under study (for descriptive statistics, see Table 7.1).

The first group of covariates were the individual characteristics of the respondent: age was used as a categorical variable (20–29, 30–35, 36–44, or 45+). The mean age of the migrants was 37, and the mean age of the non-migrants was 38. The women's educational background was captured by the highest level of education at the time of the interview (no or lower-secondary degree, upper-secondary degree, or

<sup>1</sup>The analytical distinction between unemployment and inactivity depends on the level of the benefits available and the eligibility criteria for whether individuals are defined as unemployed or out of the labour force (Black et al. 2002; Autor and Duggan 2003; Rege et al. 2009; Bratsberg et al. 2010). Some scholars have claimed that the distinction is vague and is strictly dependent on the institutional setting defined by the policies (Atkinson and Micklewright 1991). Given that our interest here is not to distinguish between different ways of not being engaged in paid work, but rather between working or not working and between different working hour arrangements, the categories unemployment and inactivity were combined.

<sup>2</sup>All of the analyses were performed using the software Stata14 and the Stata-package SPost13 (Long and Freese 2014).


**Table 7.1** Sample distribution, by migrant status, migrant generation and country of origin (column %)

Source: SLFS 2008 (N = 6814). Western Europe: Andorra, Austria, Belgium, Denmark, Finland, France, Germany, Great Britain, Ireland, Lietchtenstein, Luxemburg, Monaco, The Netherlands, Sweden; Eastern Europe: Albania, Bosnia, Croatia, Hungary, Kosovo, Makedonia, Montenegro, Poland, Romania, Serbia, Slovenia, Yugoslavia. Southern Europe: Chypre, Greece, Italy, Malta, Spain, Portugal

tertiary degree). On average, the migrants had less education than the non-migrants; 94% of non-migrants, but only 75% of migrants had at least an upper-secondary degree. In terms of their marital status, the composition of the immigrant group was rather similar to that of the non-migrant group, with the majority of the women being married (about 80%). The percentages of divorced women were rather similar as well (about 10%). Only about 1% of the mothers in our sample were widowed; this suggests that most of the women who were lone mothers were divorced or had never been married, while very few were widows (because the amount of variation was small, marital status was not used in our multivariate analyses).

The second set of control variables referred to the children. These variables were a dummy for the presence of a child between zero and two years old in the household (yes/no), and a categorical variable for the number of children under age 18 in the household (1, 2, or 3+). Overall, the number of children did not significantly vary by migrant status.

#### **Results**

#### *In and Out of Employment*

We start with a descriptive overview of the dichotomous variable on labour market participation (see Fig. 7.1). First, the share of women who were out of employment was more than twice as high among the mothers in couples as among the lone mothers (27% versus 12%). This difference was significant not only in the bivariate statistics, but also in the multivariate statistics when all our control variables were added.

**Fig. 7.1** Labour market participation by migrant status and partnership status. Source: SLFS 2008 (N = 6814)

**Fig. 7.2** Predicted probabilities of being in vs. out of employment, by partnership status and **(a)** migrant status, and **(b)** country of origin. Source: SLFS 2008 (N = 6814). LM = lone mothers, MC = mothers in couples. Estimates from logit regression models controlled for education, age, number of children under age 18 in the household, presence of a 0–2-year-old child in the household. Predicted probabilities for the controls are in the Appendix (Table 7.A1)

When we examined labour market participation levels by migrant status, we found that about 28% of the migrants and 21% of the non-migrants were out of employment (see Table 7.1). Figure 7.1 shows the results for the comparison between migrants and non-migrants by the mothers' partnership status: the pattern of lower non-employment levels among lone mothers than among mothers in couples was found for both the migrants and the non-migrants. Non-employment levels were higher for the migrants than for the non-migrants in both partnership-status groups: 17% of the lone mothers and 31% of the mothers in couples were not active on the labour market. These results remained significant in the multivariate models.

Although it was not within the scope of our paper, when we looked at the labour force participation levels of the women who had never had children or whose children had left the household we found more evidence to support the hypothesis of an immigrant disadvantage: 9% of the non-migrants and 20% of the migrants were out of employment3 .

In the next step we inserted the control variables into multivariate models, and investigated the question of whether there was variation within the migrant group. Figure 7.2a displays the linear probability of being employed versus being out of employment by migrant generation and partnership status. On the one hand, even after controlling for the individual characteristics of the mothers and their children, differences by migrant status persisted. The probability of being employed was

<sup>3</sup>Results available upon request.

rather similar for the second-generation migrants and the non-migrants: for each of these groups respectively, the probability of being employed was about 77% and 76% for the mothers in couples and about 88% and 86% for lone mothers. By contrast, for the first-generation migrants about 61% among the mothers in couples and about 72% of the lone mothers were employed. The differences between the first generation and their descendants and the non-migrants were between 10 and 15 percentage points, respectively. These differences were significant, and persist after we controlled for the characteristics of the mother and her household. On the other hand, we found clear differences in employment levels by partnership status for the first and second migrant generations, as well as for the non-migrants, with lone mothers again being more likely to be in employment than mothers in couples.

It is worth highlighting that country of origin represents a potential driver of differentiation. However, the relatively small sample size did not allow us to take into consideration migrant generation and country of origin in interaction. Nevertheless, we wanted to account for such a crucial dimension and we therefore estimated similar models distinguishing the migrants by their country of origin (Fig. 7.2b). Here we found differences of up to 15% in the FLFP of the lone mothers and the mothers in couples in nearly all groups. Compared to non-migrants, the differences by country of origin were mostly insignificant. All migrant groups had similar or lower probabilities of employment than non-migrants. For women from European countries these results may not be too surprising, as the differences between Switzerland and their respective country of origin may not be very large. However, the employment probabilities of migrants from countries outside of Europe were about 20% lower than for nonmigrants. This was a highly heterogeneous group, so our interpretation of these results is tentative. One possible explanation is that the largest proportion of migrant mothers in this group come from Turkey, where the FLFP is relatively low; thus, the results may reflect cultural differences in the FLFP in the country of origin.

#### *Working Hours*

The second part of our analyses accounted for the differentiation within employment by working hours; i.e., by full-time, long part-time, and short part-time employment. We display the results separately by partnership status and migrant generation (see Fig. 7.3a, b). The first conclusion from the Fig. 7.3a and 7.3b is that the lone mothers worked more full time or long part time than the mothers in couples, independently from their migrant status.

The second conclusion concerns the comparison between the migrant generations. Each of the three groups had distinct working hour patterns. Among the nonmigrants, only 8% of the mothers in couples and 19% of the lone mothers were in full-time employment; the lowest values in the sample. Among the first generation, 23% of the mothers in couples and 35% of lone mothers worked full time. The respective numbers for the second generation were 16% and 31%. Among the mothers in couples, the likelihood of being in short or in long part-time employment was

**Fig. 7.3** Predicted probabilities for different labour market arrangements, by partnership status and migrant generation: **(a)** Mothers in couple, **(b)** Lone mothers. Source: SLFS 2008 (N = 6814). Note: Estimates from multinomial logit regression models controlling for education, age, number of children under age 18 in the household, presence of a 0–2-year-old child in the household. Predicted probabilities for the controls are displayed in the Appendix (Table 7.A2)

not statistically different. By contrast, the lone migrant mothers were the least likely to have been in short part-time employment, independently from the migrant generation to which they belonged.

Mothers in couples were the least likely to work both in full-time and long parttime employment, when the two forms of employment were added together. This difference was most pronounced among the non-migrants (almost 40%), and it was smallest among the first generation (18%). Since migrants had higher overall rates of full-time employment (63% for the first and 80% for the second generation), this reduced the differences between lone mothers and mothers in couples among migrant mothers.

Differences in working hour patterns were consistent across the migrant groups by their country of origin (Fig. 7.4a, b). The mothers in couples of each of the country groups were more likely to have been in full-time or long part-time employment than the non-migrants, with values ranging from about 39% for the "other" migrants to 54% for the southern Europeans. For lone mothers, these numbers varied from 58% among the non-European group to 88% among the southern European group. For all of the migrant groups short part time was the least common form of employment (the minimum was 14% of lone mothers from southern Europe).

#### *Effect of Control Variables*

Finally, we come back to the effects of the control variables of the mothers and their children. We assumed that any differences by migrant status and/or by partnership status would diminish or vanish when we inserted control variables into the models. In both analyses—the analysis on the binary indicator of employment versus out of employment and the analysis on the type of employment by working hours—the

**Fig. 7.4** Predicted probabilities for different labour market arrangements, by partnership status and country of origin: **(a)** Mothers in couple, **(b)** Lone mothers. Source: SLFS 2008 (N = 6814). Note: Estimates from multinomial logit regression models controlling for education, age, number of children under age 18 in the household, presence of a 0–2-year-old child in the household. Predicted probabilities for the controls are displayed in the Appendix (Table 7.A2)

control variables did not significantly change our results concerning the main effects; i.e., the differences in employment patterns remained between the lone mothers and the mothers in couples, as well as between the migrant generations and the non-migrants by country of origin. Our results on the effects of the control variables are consistent with the literature (see Tables 7.A1 and 7.A2 in the Appendix): the presence of a child up to age two decreased the probability of being employed by 11 percentage points. The more children in the household, the less likely the mother was to be engaged in paid work. The older the mother was and the higher her education was, the more likely she was to be active in the labour force (Table 7.A1).

#### **Discussion and Conclusions**

In this chapter, we drew attention to the labour market participation patterns of women in Switzerland by focusing on two risk factors: i.e., being an immigrant from abroad or a migrant descendant and lone motherhood. We assumed that these two factors contributed to higher non-employment rates and/or marginal working conditions, and formulated four working hypotheses. These hypotheses were tested using quantitative individual data from the Swiss Labour Force Survey 2008, which contained sufficient case numbers of both target groups, and allowed us to take into account the socio-demographic characteristics of the mothers and the information on their household composition in multivariate analyses.

Our first working hypothesis concerned the comparison between lone mothers and mothers in couples in terms of their labour force participation levels and employment hours. Our analyses provided mixed evidence on this issue. On the one hand, based on the assumptions derived from household specialization theory and from the (inadequate) family-work reconciliation policies in Switzerland, we had hypothesised that lone mothers would be more likely to be out of employment than mothers in couples. This hypothesis must be rejected, as we found that lone mothers were significantly more likely to be employed than mothers in couples. This finding was consistent in all of the estimations that took into account the control variables. On the other hand, we found that if lone mothers were employed, they worked longer working hours than mothers in couples. The most common forms of employment among lone mothers were full time and long part time.

We had based our second working hypothesis on the human-capital approach, and assumed that migrants would be disadvantaged on the labour market, and would thus have higher non-employment risks and longer working hours. When estimating these outcomes for all migrant mothers in total, our results confirmed both parts of this hypothesis. Regardless of their partnership status, compared to their nonmigrant counterparts migrants were significantly more likely to be out of employment, and, if active in the workforce, to be in full-time or long part-time employment.

These results confirmed our assumption that migrant lone mothers face multiple disadvantages. On the one hand, migrants are disadvantaged in terms of finding employment for a variety of reasons (e.g., their diploma or qualifications may not be recognised, they may be discriminated against by employers, or they may face limitations because of their residence permit); thus, when they find a job they may be less able to negotiate a part-time schedule that would help them reconcile their work and care responsibilities. On the other hand, a lone mother who provides the main income in the household may have few options other than to work full time, and may be less able to afford being out of employment or in a part-time job than a mother in a couple. Thus, for migrant lone mothers a double imperative seems to apply: they can either work full time if the income from work is sufficient to pay for external (expensive) care and sustain the household's costs, or they can live on social assistance if their prospects of finding gainful employment are gloomy. In the former case, the stresses associated with reconciling work and family may be greater than for mothers in couples, for whom working part time may provide a partial solution.

In further analyses we examined within-group variation among immigrants, thus providing evidence for our third hypothesis: i.e., we found that when we only considered being in or out of employment, women of the first migrant generation in particular were disadvantaged relative to non-migrants. The patterns of the second generation were not, however, significantly different from those of non-migrants. By contrast, when we analysed the type of employment, we found that the differences between first-generation migrants and non-migrants remained; i.e., that firstgeneration migrants had higher levels of non-employment and longer working hours. But the employment patterns of the second generation appeared to fall between those of the first generation and the non-migrants. These patterns were consistent in all of the models that took into account the control variables: in line with the literature, our results showed a strong immigrant disadvantage among firstgeneration migrants in particular, as well as evidence of adaptive processes and increasing integration among migrant descendants. But we also found that when

two risk factors were combined—i.e., lone motherhood and migrant status—an additive disadvantage on the labour market emerged.

We also investigated within-group variation by country of origin. The results produced here were conclusive in that they showed that origin did not matter in addition to the migrant status, because the patterns were consistent across almost all of the groups. Ideally, we would have taken into account the migrant generation and the country of origin simultaneously, but the small sample size of these two risk groups did not allow us to do so. We therefore do not want to offer an interpretation of these results, other than to observe that the rather small differences found between the mostly European countries of origin may be reflected in the small amount of variation found among the immigrant groups as well. The differences between the European immigrants and the group of "other" migrants were slightly larger, but because this group was so heterogeneous we cannot draw any reliable conclusions here. Overall, it seems that migrant generation and partnership status are the factors that matter most for the labour market participation levels of immigrant women in Switzerland.

At this point, however, we return briefly to the fourth working hypothesis on the effect of compositional differences by the socio-demographic characteristics of the mothers and their children. Inserting these control variables did not significantly change our results: both the lone motherhood disadvantage and the migrant disadvantage remained prominent in our analyses. These results suggest that future research on women's labour force participation should take into account withingroup variation among mothers by their partnership status, as well as among migrants by accounting for migrant generation and country of origin.

Thus, our findings suggest that the common assumption that social protection measures tend to discourage LFP among lone mothers with unfavourable labour market prospects should be revised (Gangl 2006; Luijkx and Wolbers 2009; Dieckhoff 2011). Relative to the mothers in couples, the lone mothers in our sample were more likely to be in the labour market and to be working long hours. An open question for future research to investigate is the extent to which employment and working hour patterns relate differently to the health risks faced by both parents and children depending on whether they are migrants. For lone mothers, the positive association between employment and health may be weakened by the double burden of being the main earner and the main caregiver: flexible working hour arrangements could counterbalance the potential additional stress of being employed in the absence of work-family balance policies (see Struffolino et al. 2016 for a review). However, migrants may have limited access to such arrangements due to their segregation in specific labour market sectors.

In addition to other problematic issues connected to the status of migrants (e.g., insufficient language skills, below-average education and job qualifications, religious and non-religious forms of discrimination), difficulties in accessing child care may be one of the reasons why first-generation migrant mothers have a particularly high unemployment risk. They may have smaller support networks for child care than non-migrant lone mothers. Previous research has demonstrated that care of grandchildren is an important form of intergenerational support provided by the older generation to the younger generation. Especially when public child care is scarce or/and expensive, this type of support may be necessary for mothers to return to the labour market (Wheelock and Jones [2002] for England). Hank and Buber (2007) showed that child care provided by grandparents can reduce the costs of (women's) labour market engagement. This financial consideration may particularly relevant for migrant mothers, who often work in low-wage jobs. Thus, compared to non-migrant women, who can rely on help with child care from their own parents or their partner, migrant women, who often lack a supportive social network in the country of destination, may face additional challenges in returning to work, especially if they are not in a couple.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **Appendix**


**Table 7.A1** Predicted probabilities (PP) for being employed vs. not being in employment. 95% confidence intervals (CI)

Note: Estimates from binary logit regression models. Model (a) controls for migrant status, model (b) controls for country of origin. SLFS 2008 (N = 6814).

**Table 7.A2** Predicted probabilities (PP) for the control variables for not being in employment or being in full-time or long or short part-time employment, by migrant status and partnership status (Fig. 7.3), and by country of origin and partnership status (Fig. 7.4). 95% confidence intervals (CI)



**Table 7.A2** (continued)

Note: Estimates from multinomial logit regression models. SLFS 2008 (N = 6814)

#### **References**


Bühler, E. (2002). *Atlas suisse des femmes et de l'égalité*. Zürich: Seismo.


**Nadja Milewski** is Assistant Professor for Demography at the Institute for Sociology and Demography at the University of Rostock, Germany. Her research centres on demographic behaviour and life courses of international migrants, i.e., fertility and family dynamics, labour force participation and gender roles, as well as health and morbidity.

**Emanuela Struffolino** Emanuela Struffolino is senior researcher at the "Demography and Inequality" research group at the WZB Berlin Social Science Center and at the chair of Microsociology at the Humboldt University of Berlin, Germany. She obtained her PhD in Sociology at the University of Milano-Bicocca, Italy, and worked as research fellow at the Swiss NCCR program "LIVES overcoming vulnerability: life course perspectives". Her research interests include life-course sociology, gender inequalities in the labour market, social stratification, and methods for longitudinal data.

**Laura Bernardi** is Professor of Life Course Demography and Sociology at the Faculty of Social and Political Sciences at the University of Lausanne, Switzerland, and Deputy Director of the Swiss National Center for Competence in Research LIVES. Her major research areas are fertility, family demography, migration and life course. She has published in major sociological and demographic journals on topics like fertility intentions and behaviour, social network influences on family development, intergenerational transmission of fertility and family formation, lone parenthood and health, migration and family formation. She is also co-editing the Springer series on Life Course and Social Policies.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Part III Income and Poverty Among Lone Parents**

## **Chapter 8 'Only a Husband Away from Poverty'? Lone Mothers' Poverty Risks in a European Comparison**

**Sabine Hübgen**

#### **Introduction**

In the EU at-risk-of-poverty rates1 have been quite stable over the last decade (Eurostat 2014). However, some social groups – among them lone mothers – are disproportionately affected by income poverty. Sociologists explain this phenomenon largely referring to the term of "new social risks" meaning that welfare state institutions did not adjust properly to major demographic and social changes. Thereby, the case of lone mothers appears to be crucial, as their prevalence increased tremendously in the course of the pluralization of family forms over the last decades (European Commission 2007; Jaehrling et al. 2011). Despite this overall trend, we observe substantive variations in lone mothers' poverty risks across the EU member states (Brady and Burroway 2012; Christopher 2002; Lelkes and Zólyomi 2008; Misra et al. 2007): The at-risk-of-poverty rates vary between 13% in Denmark and 49% in Luxembourg (see Table 8.2). Moreover, the at-risk-of-poverty rates for lone mothers do not necessarily reflect a country's overall at-risk-of-poverty rate. In the Czech Republic, for instance, the overall at-risk-of-poverty rate is comparatively low, whereas the at-risk-of-poverty rate for lone mothers ranks among the highest. Similarly, while German lone mothers' poverty risk is twice as high as that of the overall population; Danish lone mothers face even lower poverty risks than the overall population. These descriptive numbers already reveal that an absent partner *per se* cannot account for this striking variation in lone mothers' poverty risks. On the contrary, it appears rather to be an indicator for to which extent a lone mother is capable of providing a decent standard of living to herself and her children in a specific country (Hobson 1994:171). From previous research we know that besides

S. Hübgen (\*)

<sup>1</sup>This term refers to the EU's At-risk-of-poverty-rate at the 60% threshold.

WZB Berlin Social Science Center, Berlin, Germany e-mail: sabine.huebgen@wzb.eu

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_8

individual factors (e.g. age, work status, education, the age and number of kids) also institutional contexts play an important role for understanding lone mothers' poverty risks (Brady and Burroway 2012; Huber et al. 2009; Misra et al. 2007, 2012). Though, from a theoretical point of view the mechanisms of these associations remain rather unclear.

Hence, this chapter aims to shed some more light on the influence of institutional configurations on lone mothers' poverty risks. Institutional arrangements shape the individual's scope of action in both structural and normative regards. With respect to lone mothers' poverty risks, the labor market on the one hand and the welfare state on the other constitute the two key institutions: they determine to what extent lone mothers have either access to labor income and/or social transfers. On theoretical grounds this chapter builds upon the prolific work of gender-sensitive welfare state research. One major argument is that high at-risk-of-poverty rates for lone mothers can be interpreted as an indicator for high gender inequalities because those mothers seem to be unable to making ends meet without a male partner in the household (Hobson 1994). This assumption will be tested empirically focusing on gender inequalities in the labor market and the welfare state. Furthermore, I want to investigate the interplay of these institutional configurations with relevant individual characteristics for poverty outcomes. Data stem from multiple waves (2009– 2012) of the *EU Statistics on Living and Income Conditions* (EU-SILC). Addressing the hierarchical data structure multi-level models are estimated in the multivariate analyses.

#### **Theory and Hypotheses**

Over the last decades the proportion of lone parents on all families has been visibly increasing (European Commission 2007). In some European countries lone parents – but mainly lone mothers – represent between 20% and 25% of all families (Jaehrling et al. 2011; Kiernan et al. 2011). However, lone motherhood is still largely considered either as predictor for overall poverty (Brady et al. 2009; Lelkes and Zólyomi 2008; Vandecasteele 2011) or studies emphasize the negative consequences for children growing up in lone parent families (Brooks-Gunn et al. 2002; Kiernan et al. 2011; Sawhill 2003, 2014). There also exist some rich descriptive case studies which focus on lone mothers' family life, employment patterns, welfare reception sometimes also taking lone mothers' poverty risks into account (Bahle et al. 2013; Fux 2011; Heimer et al. 2009; Jaehrling et al. 2011; Ott et al. 2011; Zagel 2014). Other studies inspired by the life course perspective refer implicitly to lone mothers investigating the economic consequences of risky life events – family break-ups among them (Andreß et al. 2006; Kohler et al. 2012; also see Harkness and Mortelmans and Defever in this book, Radenacker 2011). Comparative studies which explicitly attempt to explain lone mothers' poverty risks are still rare. The poverty literature in general differentiates between individual and structural causes of becoming poor. Well established individual poverty predictors like age, age and number of children, marital status, level of education, employment status and working hours are also crucial for lone mothers' poverty risks (Brady and Burroway 2012; European Commission 2007; Fux 2011; Jaehrling et al. 2011; Lelkes and Zólyomi 2008; Misra et al. 2007, 2012; Ott et al. 2011). This is particularly true as there is no (male) partner on the household level who could level off (at least some of) the disadvantages. Regarding structural explanations of lone mothers' poverty risks, in previous studies particularly characteristics of the labor market and the welfare state turned out to be important.

#### *The Labor Market*

For instance, being employed would generally reduce a lone mother's poverty risk significantly. However, her employment status and working hours are not only dependent on individual qualifications and decisions, but also on characteristics of the labor market. These characteristics are embedded in and shaped by a specific historical, political and normative context. In countries with a long tradition of female labor force participation it is easier for (lone) mothers to get access to paid work and longer hours than in countries with a strong tradition of the 'male breadwinner-model'. In fact, nowadays the majority of women in Europe are employed, but in many countries we can still observe a clearly gendered responsibility for childcare. These gendered norms do not only lead to the so called 'double burden' for mothers, but also to a 'motherhood penalty' – i.e. economic disadvantages in career development and earnings compared to childless women or to fathers (Benard and Correll 2010; Gangl and Ziefle 2009; Grunow et al. 2011).

Hence, lone mothers' poverty risks might vary between countries dependent on the structural characteristics of the labor market – and particularly gender inequalities. In a first step existing gender inequalities in the labor market have an impact on the *access* to paid work for women and especially for mothers. Consequently, in some countries the access to labor income is much more equal than in others. In a second step, countries might differ regarding the degree of *occupational sex segregation* – in which types of occupations, positions and working conditions women can find a job (Grunow et al. 2011; Mandel and Semyonov 2006). Thereby, 'femaletypical' occupations are often characterized by relatively low earnings and working hours and meager career opportunities (Bardasi and Gornick 2008). This in turn results in (long-term) poverty risks. Similarly, Jaehrling et al. (2011: 53) point to the fact that part time work often stands either for a "revolving door" into unemployment or for a "dead end street" cumulating all the corresponding disadvantages over time. On this basis I want to test empirically whether existing gender inequalities in the labor market have an effect on lone mothers' poverty risks: The more pronounced gender inequalities – referring to the access, working hours and earnings – exist in the labor market, the higher are lone mothers' poverty risks in the respective country (*Hypothesis 1a*).

#### *The Welfare State*

Besides the role of the labor market, many studies found welfare state generosity to be a crucial predictor for poverty in general (Brady et al. 2009; Esping-Andersen 1990; Huber et al. 2009; Misra et al. 2007; Smeeding 2005). Some researchers criticize the concept of generosity – largely measured as social spending –, as it does not take into account the political objectives and the quality of the programs (Misra et al. 2007: 806). Hence, more recent studies investigate the influence of specific policies or laws on poverty. As an example, Brady and Burroway (2012: 738) show that lone mothers' poverty risks depend on whether a *welfare state's organization* is based on the principle of universalism and citizenship versus a strong emphasis on 'targeting'. Furthermore, feminist scholars argue that *social rights and eligibility rules* are not gender-neutral, so that welfare states reproduce gender inequalities (Orloff 2009; Sainsbury 1999). Clearly, this can be closely linked to the organizational principle of welfare states because social rights based on citizenship leave only little room for gender discrimination. On the contrary, in countries like Germany with a strong focus on earnings-related social insurances, persons without a life-long standard employment relationship are systematically disadvantaged. Hereby particularly mothers are affected when they interrupt their careers right after the birth of a child. Similarly, eligibility rules also can vary by marital status providing certain privileges to married couples. This is also very relevant for lone mothers because in some countries divorced or widowed lone mothers have access to more generous social benefits than never married lone mothers (Hobson 1994: 182).

So far, the impact of gender-specific eligibility rules has not been tested empirically. Beyond that, existing *work-family-policies* are particularly important for lone mothers as they shape their possibilities to reconcile paid labor work and unpaid care work. Admittedly, this policy area is strongly affected by existing gender norms and role models (Pfau-Effinger 2004, 2005). This can be perfectly illustrated by the example of public childcare: While in most European countries care for children from the age of three is largely provided, in some countries care for the younger is strongly debated and the coverage is consequently still very low. The results by Misra et al. (2007, 2012) indicate that generous family benefits on the one hand and a broad supply of public childcare on the other reduce lone mothers' poverty risks substantively. In contrast, the effects of paternity leave policies are mixed: A generous replacement rate can reduce poverty risks right after birth, but it can have the reversed effect if long leaves are granted (Aisenbrey et al. 2009, ebd: 2007, Jaehrling et al. 2011, Mandel and Semyonov 2006). Building on previous literature I assume that the more pronounced gender inequalities exist in the welfare state – in terms of gender-specific eligibility rules and insufficient work-family policies –, the higher lone mothers' poverty risks in the respective country *(Hypothesis 1b)*.

#### *The Interplay of Institutional Configurations and Individual Characteristics*

One shortcoming of previous research consists in that the impact of individual characteristics and institutional configurations on lone mothers' poverty risks have been investigated quite independently of each other. As a consequence, the exact interplay of individual characteristics in specific country contexts is still unclear. This chapter aims at closing this gap at least partly: With regard to lone mothers' employment I argue that its poverty-reducing impact is not only dependent on individual characteristics like qualifications, but also on the extent of existing gender inequalities on the labor market. More explicitly, the more pronounced gender inequalities are regarding working hours and earnings, the higher are lone mothers' poverty risks despite of being employed (*Hypothesis 2*).

Similarly, the poverty-enhancing effect of children might not be identical across Europe due to varying degrees of gender inequalities in terms of work-family policies. In countries where care is clearly ascribed to the family sphere – thus to the mothers –, welfare states will only provide rudimentary public childcare. Hence, I assume that the more traditional work-family- policies are in a country, the greater young children's poverty-enhancing impact on lone mothers (*Hypothesis 3*).

#### **Data, Measures and Methods**

#### *Data & Sample*

Individual data are pooled together from four EU-SILC waves (2009–2012). In doing so, a reasonable number of lone mothers per country is obtained. Bulgaria, Croatia, Cyprus, Malta and Romania are excluded either due to low case numbers (<400) or major data issues. The final country sample covers 25 European countries, country-level characteristics stem from the Eurostat data base. Lone mothers are defined as follows2 : A woman who lives only with her dependent child(ren). The term 'dependent' here include all children below the age of 18 and children up to the age of 24 when either in education or unemployment. Mothers identified as 'livingapart-together' with a husband are excluded from the sample, as we can assume them to form an economic unit. Moreover, the sample is restricted to lone mothers aged 18–59 years. The final sample counts 28,738 lone mothers from 25 countries.

<sup>2</sup>This 'narrow' definition excludes for example lone mothers living with their parents or with adult children. However, in all sample countries (except Poland and Slovakia) the vast majority of lone mothers live only together with their dependent children. Moreover, this chapter focuses on the relationship between institutional gender inequalities and poverty. Therefore, it is preferable to not confound these mechanisms with others stemming from different HH constellations.

#### *Measures*

Following the official EU definition, a lone mother is defined as *being at risk of poverty* if her net household income sums up to less than 60% of a country's respective median of the national net equivalent household income. As I am interested in the influence of national differences regarding institutional configurations, a relative definition of poverty seems to be adequate. The dependent variable is coded as a dummy, poor (=1) or not poor (=0).

Lone mothers' individual characteristics are measured as follows: self-defined3 *employment status and working hours* are combined in 4 dummy variables: full time employed, part time employed, unemployed and inactive. Further, the *total number of children* (1; 2; 3 or more) and the presence of *very young children* (0–2 years) are included. Besides, models will control for level of education (high, medium, low), marital status (never married, divorced, widowed) and mother's age (in years).

Existing gender inequalities in the labor market and the welfare state are captured by several indicators (see Table 8.1 for an overview: The gender-specific access to paid labor is measured as the *Female Labor Force Participation* Rate, whereas two indicators of occupational sex segregation are included: the *Gender Pay Gap* and the *Female Full Time Employment Rate* as a share of the female labor force. Furthermore, work-family policies are represented by the *Provision of Childcare* (childcare usage rate for children up to the age of 3) and *Family Benefit Generosity* (in PPPs per capita). The degree of *Gender-specific Eligibility* rules in the welfare state is measured as the share of social contributions on total social protection receipt. Finally, I will control for the overall working age at-risk-of-poverty rate because on the one hand it nets out the specific poverty risks of lone mothers from general poverty risks. On the other hand it serves as an indirect indicator for overall welfare state effectivity across countries.

#### *Analytical Strategy*

The empirical analyses start out with a short descriptive overview on lone mothers' at-risk-of-poverty rates across the 25 countries in comparison to the overall population and partnered mothers. The following sections will analyze to what extent individual and institutional factors and finally the interplay of both can account for this variation in lone mothers' poverty risks. First, lone mothers' social composition across countries will be presented. It might be the case that in some countries particularly socially disadvantaged women become lone mothers, whereas in other countries this family form is less negatively selected. Thus, some of the country

<sup>3</sup>A self-defined status seems to be superior because there are no universal thresholds across occupations and countries. However, 'full time' is usually reported when working for at least 30 hours per week (Eurostat 2009: 177).


**Table 8.1** Overview of gender inequality indicators and the overall at-risk-of-poverty rate across 25 countries

Source: Eurostat Database based on data from European Labour Force Survey and EU-SILC Note: Displayed values are averaged across the years 2009–2012. *AROP=* At-risk-of-Poverty Rate

variation in lone mothers' poverty risks could be a result of different selection processes leading to lone parenthood. Then, institutional configurations focusing on gender inequalities will be illustrated and linked to lone mothers' at-risk-of-poverty rates across countries.

Building on these descriptive parts, several multi-level models are employed for hypothesis-testing. Multi-level procedures allow the simultaneous modelling of individual and higher level factors on an individual criterion variable addressing some major statistical challenges regarding mixed level analysis (e.g. correct standard errors). Furthermore, the interplay of both individual and context characteristics can be modelled with so called 'Cross-level interactions' in Random Slope Models. Random Slope Models allow specific coefficients to vary across countries. Cross-level interactions prove whether this added variance can be 'captured' by country-level characteristics – in this case by gender inequalities in the labor market and the welfare state. This procedure is used for testing Hypothesis 2 and 3. Despite the binary outcome variable (poor/not poor), the models are specified as linear probability models with robust standard errors. Following Mood (2010) this procedure addresses the problem of counterintuitive and misleading interpretation stemming from Logits or Odds Ratios. Furthermore, logistic multi-level models suffer repeatedly from convergence problems, especially when running more complex Random Coefficient Models.

#### **Descriptive Findings**

#### *Lone Mothers' At-Risk-of-Poverty Rates & Poverty Ratios*

An overview of lone mothers' at-risk-of-poverty (AROP) rates is presented in Table 8.2 compared to overall and partnered mothers' AROP rates and the respective poverty ratios. Almost one third of the lone mothers (31%) in this country sample are at risk. As mentioned above, lone mothers' AROP rate is highest in Luxembourg (49%) and lowest in Denmark (13%). Countries fit only partly into well-known welfare state typologies: In three Scandinavian countries (DK, FI, NO) comparatively few lone mothers face poverty risks, whereas the Swedish AROP rate for lone mothers lies above average AROP rates and resembles more the British one. Similarly, among the high poverty countries two 'conservative' welfare states (DE, LU) cluster together with Lithuania, Latvia and Spain. As expected the so called 'liberal' welfare states of UK and Ireland show AROP rates for lone mothers above average, but they are not among the highest.

Table 8.2 also provides the ratios of lone mothers' poverty risks compared to overall poverty risks and that of partnered mothers. These poverty ratios also point in the direction that lone mothers' poverty underlies specific mechanisms: In almost all countries lone mothers face higher risks of poverty than the overall working age population. In the Czech Republic and Luxembourg for example lone mothers' AROP rate is 2.5 times higher. Even more pronounced are the differences in AROP rates between lone and partnered mothers. On average, lone mothers face 2.8 times higher poverty risks than their partnered equivalents. This poverty ratio is particularly high (>4) in the Czech Republic, Norway, Germany and Sweden. Maybe surprisingly, in the Southern European countries and Poland AROP rates for lone and partnered mothers are quite similar. It is also in these countries where other living arrangements are quite common among lone mothers.

So it might be the case that the sample lone mothers for these countries form a rather specific group which can afford to live without other family members and therefore do not differ significantly from partnered mothers.


**Table 8.2** At-risk-of-poverty rates and ratios for lone and partnered mothers and the overall population

Notes: Weighted At-risk of-poverty rates; N = 25. Countries sorted by Lone Mothers' AROP Rate (descending). *AROP*= At-risk-of-Poverty rate

#### *Social Composition and Lone Mothers' Poverty Risks*

A first potential explanation of this great variation in lone mothers' poverty risks across countries could be different selection4 processes leading to lone motherhood in different countries. Table 8.3 in the Appendix presents the share of 'risky' characteristics among lone mothers across countries. The column 'Number of risks'

<sup>4</sup>The most preferable modelling strategy would be a Heckman selection correction (Heckman 1979). In practice however, it is quite challenging to find a suitable instrument variable which is correlated with lone motherhood, but not with poverty outcomes. Therefore, I employ a more descriptive approach to detect the role of lone mothers' social composition for their poverty risks.


Lone mothers' social composition and number of above average prevalent risks across 25 European countries (in %)

176

average sums up in how many of the seven considered characteristics a country scores above average. Countries are sorted by this number of risks. For example Belgian lone mothers show in six out of seven characteristics comparatively high proportions: They are distinctively more often either inactive or unemployed, low educated, never married and live with numerous and/or young children. On the contrary, in Slovakia lone mothers on average tend to be older, employed, well educated, divorced and live with only few and older children. Accordingly, we would expect that countries with a high proportion of 'risky' characteristics among lone mothers also show higher at-risk-of-poverty rates. This holds true for some of the sample countries like Belgium, Luxembourg, Ireland and the UK. Inversely, in Finland, Hungary, Poland, Slovakia and Switzerland comparatively low proportions of risky characteristics are also associated with comparatively low at-risk-of-poverty rates for lone mothers.

However, for more than half of the sample the pattern is either not that clear or even in reverse: In Germany and the Czech Republic for example lone mothers are strongly at risk of poverty although their social composition is rather favorable. Except for the comparatively high unemployment rates for lone mothers, which might drive the high poverty risks. In contrast, Iceland and Norway, among those countries with the lowest at-risk-of-poverty rates for lone mothers, show a rather negatively selected social composition: One fourth and one third of all lone mothers is low educated. Similarly, the share of never married lone mothers ranks among the highest. Moreover, in Iceland and Norway young and inactive lone mothers with numerous (Iceland) and young children are more common than in other European countries. Even Danish lone mothers who face by far the lowest poverty risks show average rates of inactivity and unemployment and rather high shares of low education and out-of-wedlock births. Hence, social composition might account for some variation in lone mothers' poverty risks, but the same individual characteristics seem to translate into poverty risks only in some countries. This can be counted as an indicator for the importance of institutional configurations which shape individual risks and opportunities.

#### *Institutional Configurations*

This chapter stresses the role of gender inequalities in the labor market and the welfare state for lone mothers' poverty risks. Figure 8.1a–f provides some insights into the countries' institutional configurations in form of bivariate correlations. In order to relate these configurations to lone mothers' poverty risks country markers indicate whether a country's AROP rate for lone mothers lies above (filled) or below (hollow) the average. Figure 8.1a presents the positive and moderate correlation between the Female Labor Force Participation Rate and the Gender Pay Gap. Thus, in countries where most women work, they also tend to face comparatively higher pay disadvantages. The countries with the lowest at-risk-of-poverty rates for lone mothers also have comparatively high female labor force participation rates, but

**Fig. 8.1** Correlations of gender inequalities indicators (Note: **Filled** marker symbols indicate lone mothers' AROP Rate **above average** (e.g. DE) and *hollow* markers *below* average (e.g. NL); N = 25)

rather show a medium degree of Gender Pay Gap. In some Eastern European countries (HU, PL; SI, SK) lone mothers face comparatively low poverty risks, although only around 50% of the women in these countries are integrated into the labor market. Figure 8.1b reveals that most of those women who actually are in the labor market work full time which might be a reason for the low at-risk-of-poverty rates in those respective countries. In general, lone mothers are the least affected by poverty in countries where broad access to paid labor for women is combined with good opportunities for working full time. In contrast, in countries where a high female labor force is combined with low full time employment rates lone mothers' poverty risks tend to be comparatively high (i.e. DE, SE, UK). Similarly, a high female full time rate which is restricted to only a small share of the female population is associated with high at-risk-of-poverty rates for lone mothers.

Female employment is generally supported by the provision of formal childcare which helps mothers to reconcile paid work and care work. Surprisingly, female full time employment is negatively correlated with child care usage (Fig. 8.1c). In most countries where at least one third of the parents of very young children use formal childcare, full time employment rates for women lie below the average (except for ES, PT, SI). This negative association is mainly driven by East European countries where female full time employment is extremely high and childcare usage for children below the age of three is very low. This can be partly attributed to comparatively long parental leave durations in those countries so that most mothers would care at home for their young children until they enter kinder garden. Additionally, in some Eastern European countries public childcare has a bad reputation which encourages mothers even more to make full use of the parental leave entitlement (Heinen and Wator 2006: 205 for Poland, Saxonberg and Szelewa 2007: 359 for the Czech Republic and Poland). This negative correlation of female full time employment and childcare usage is also driven by the reversed case of the Netherlands: Although childcare usage is comparatively high, only few women work full time. This is not surprising as the short opening hours of most Dutch day care centers only enable part time work. As a consequence, the impact of childcare for lone mothers' poverty risks is not straightforward: In some countries (like Scandinavia) it enables lone mothers with young children to reconcile care and full time work which then prevents poverty.

However, in other countries (e.g. UK) childcare is rather market-based and therefore expensive which might eventually enhance poverty risks. Accordingly, in countries where motherhood is rewarded quite generously, female full time employment tends to be rather low (Fig. 8.1d). The cases of Luxembourg, Germany and Ireland illustrate well that even very generous family benefits do not seem to help lone mothers much to avoid poverty. Moreover, there is no strong correlation between the generosity of family benefits and the provision of formal childcare. While some countries clearly emphasize either family benefits (AT, CH, DE, IE) or formal childcare (ES, NL, PT, SE), most of the countries provide a mix of both (Fig. 8.1e). Furthermore, Fig. 8.1f shows the association of the provision of childcare and the organization principle of welfare states. Welfare states that rely to a great extent on social contributions often do not promote mothers' employment through public childcare. This is especially the case in many Eastern European countries. In Denmark, Iceland, Norway and Portugal the combination of a low share of social contributions and high formal childcare provision comes along with low poverty risks for lone mothers.

Summing-up the descriptive results so far: Lone mothers seem to face comparatively low poverty risks in countries where formal childcare is largely provided in combination with a more universal welfare state. Formal childcare provision enables lone mothers to reconcile care and labor. Therefore, it is important that lone mothers get broad access to the labor market and long hours. However, these patterns should only be seen as tendencies and there are always exceptions. Hence, the following multivariate analyses will shed some more profound light on the associations of structural gender inequalities and lone mothers' poverty risks.

#### **Main Findings from Multi-level Analysis**

#### *Do Gender Inequalities Influence Lone Mothers' Poverty Risks?*

Table 8.4 presents the results from different Random Intercept Models on lone mothers' poverty. The Random Intercept Only Model shows that the average probability of being poor even for a 41 year old full time employed, highly educated and divorced lone mother living with only 1 child older than two lies at 30%. Then, individual characteristics are included in the model (column 2). All coefficients point into the expected direction. The effects of marital status, number and age of children are quite small, though statistically significant. The individual effects remain robust when introducing country-level indicator of gender inequalities. The third model shows the coefficients of the three labor market indicators (Female Labor Participation, Female Full time Employment and Gender Pay Gap): They are all very small and only that of female full time employment is statistically significant. However, the reduction in the ICC indicates that some of the variance on the country level is captured by these variables. The Female Full Time Employment Rate (FFTER) seems to render a positive – thus enhancing – effect on lone mothers' poverty risks: When the FFTER increases by 1 standard deviation, then lone mothers' poverty risks increase by 4%. This rather unintuitive result will be analyzed in further detail in the next subsection. Next, a model with the three welfare state indicators is specified. While the two work-family policy indicators (childcare usage and family benefit generosity) are close to zero and insignificant, an increasing share of social contributions seems to aggravate lone mothers' poverty risks. The final model combines the two most relevant context indicators from the previous models. Both coefficients for female full time employment and share of social contributions decrease slightly, but show a robust poverty-enhancing effect for lone mothers. Deriving from this model there is no clear empirical evidence for hypothesis 1a, but there is some for 1b regarding the organization principle of welfare states.


**Table 8.4** Estimates from multi-level linear probability models with random intercepts

Notes: Raw coefficients on individual level; age centered. Z-standardized coefficients at country level

+ *p* < 0.10, \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.001

#### *Do Gender Inequalities Affect the Effects of Individual Characteristics on Poverty?*

The so far specified models treated all individual characteristics as identical across the 25 countries (fixed effects). However, I argued that these effects might rather vary across countries in the case of employment and presence of young children. And indeed, separate regression models for each country show a clear variation in effect size for full time and part time employment and to a smaller degree for very young children in the household (Figs. 8.A1 and 8.A2 in the Appendix). In order to address and exploit this variation two Random Slope models for each of the two variables are specified. These models allow the coefficient of full time employment (and respectively the presence of a child below the age of three) to vary across countries. Cross-level interactions then prove whether or not (some of) this variation can be 'explained' by gender inequality indicators. Accordingly, Table 8.5 contains 2 models per random coefficient, one with the individual characteristics and the main context indicator and another adding the respective cross-level interaction. The first column presents the Random Coefficient Model for *full time employment* with individual characteristics and the Female Full Time Employment Rate. As expected adding a random coefficient increases on the one hand the random intercept and on the other hand the Intra-class correlation5 (in comparison to the corresponding Random Intercept Model; Table 8.4, column 2). The Random Coefficient can be interpreted as the variance of full time employment across countries. This variance lies at .005 which appears rather small. This value corresponds to a standard deviation of .007. This means that in 95% of the cases the effect of full time employment varies between −.59 and −.31.

The next model (column 2) proves whether this variation might be a consequence of a differing FFTER across countries. And indeed, the coefficient for the crosslevel interaction of FFTER and lone mothers' full time employment is statistically significant and negative, although the overall model fit is not significantly improved. In countries with a high FFTER the poverty-reducing effect of full time employment compared to other employment categories is stronger than in countries with a low FFTER. This result is illustrated in Fig. 8.2. It shows the predicted probabilities of being at risk of poverty for lone mothers when working full time in contrast to working part time, being unemployed or inactive at different levels of Female Full Time Employment on the country level. With increasing Female Full Time Employment this gap grows wider: Within a country with a high share of Female Full Time Employment (e.g. HU, SK) the poverty-reducing effect of being full time employed compared to all other employment categories is larger than in countries with a comparatively low share of Female Full Time Employment. This finding goes in line with *Hypothesis 2*. Moreover, the figure also indicates that the poverty-

<sup>5</sup>The so called ICC indicates in the RIO Model the amount of variance which can be explained on level 2. In subsequent models it serves as an indicator for how much of this variance remains after introducing level 2 indicators.

enhancing effect of higher FFTER remains robust: In countries where the FFTER is low, lone mothers' predicted probabilities of being at risk are lower than in countries with a high FFTER. This counterintuitive result becomes clearer when keeping Fig. 8.1b, c in mind. High FFTER countries are mainly Eastern European countries with either underdeveloped or low quality public childcare. Thus, it is particularly difficult for lone mothers to reconcile labor and care responsibilities. Furthermore, in most countries where full time employment among women is low, the overall female labor force participation is high, thus the majority of women work part time. A potential interpretation is that in countries where female employment is common women can have good career opportunities and have access to well-paid jobs. This results in comparatively lower predicted probabilities for being poor as a full time working lone mother. In contrast, in countries where most of the women in the labor force work full time, but the female labor force as a whole comprises only up to 50% of the working age women, lone mothers might find only poorly paid full time jobs which protect them less from being poor. Overall, being full time employed reduces lone mothers' predicted probabilities of being poor compared to other employment status.

Column 3 and 4 in Table 8.5 represent the corresponding Random Coefficient Models for living with a very young Child. The fixed coefficient is close to zero and statistically insignificant. But, it is worthwhile to have a look at the country-specific coefficients for living with a very young child. Admittedly, effect sizes are rather small in most countries, but in some countries the sign of the coefficient is positive and in others negative. This variation could result in a quasi-zero association when pooling all countries together. Therefore, it is interesting to allow this coefficient to vary across countries. The variance lies at .10 which means that the fixed coefficients vary between −.21 and +.21. However, the Cross-level interaction model shows no empirical support for *Hypothesis 3*, as the interaction term is not statistically significant and does not show the expected negative sign. Further, both the BIC and the Likelihood-Ratio Test indicate a decline in model fit when adding the cross-level interaction.

#### **Conclusion**

This chapter aimed to shed some light on the variation in lone mothers' poverty risks across Europe. Descriptive in nature this chapter does not provide any causal inferences or dynamics of lone motherhood and poverty. Instead the idea was rather to introduce the framework of gender inequalities into the research of lone mothers' poverty risks and to model some of the main theoretical mechanisms of the institution-individual-nexus.

In order to analyze this phenomenon I compared lone mothers' poverty risks, individual characteristics and institutional configurations across 25 European countries. And the empirical analyses show that lone mothers' poverty risks follow specific mechanisms: Lone mothers' AROP rates do neither correspond to overall


**Table 8.5** Estimates from multi-level linear probability models with random slopes and crosslevel interactions

Notes: Raw coefficients on individual level; age centered. Z-standardized coefficients at country level

+ *p* < 0.10, \* *p* < 0.05, \*\* *p* < 0.01, \*\*\* *p* < 0.00

**Fig. 8.2** Predicted probabilities of being at risk of poverty for lone mothers by employment status at varying degrees of female full time employment rates (Note: Confidence intervals are displayed at the 95%-level)

at-risk-of-poverty rates nor to at-risk-of-poverty rates for partnered mothers. Descriptive analyses could also show no clear evidence for composition effects explaining the striking variation in lone mothers' poverty risks. Therefore, the focus was laid on the role of gender inequalities in the labor market and the welfare states for lone mothers' poverty risks. Overall, existing gender inequalities appear to account partly for s the variation in lone mothers' poverty risks across countries (Hypothesis 1(a) and b). Furthermore, this chapter contributes to the existing literature on lone mothers' poverty risks as it explicitly models the interplay of institutional factors and individual characteristics. And indeed, there is some empirical evidence that the poverty-reducing effect of full time employment is not identical across countries, but rather dependent on the respective employment regimes. Including Eastern European countries into the analyses also challenges the broadly acknowledged positive linear association of childcare provision and female full time employment and how both indicators relate to lone mothers' poverty risks. A high FFTER does not necessarily correspond with low poverty risks for lone mothers like in Estonia, Latvia or the Czech Republic. Either lone mothers don't get access to those jobs as they cannot find good and feasible childcare or they often end up in low paid full time jobs which cannot prevent them from being at risk.

Though, this chapter finds no clear interaction effect of childcare provision and the poverty-enhancing effect of young children in the household. In part this might be due to measurement problems as the childcare indicator includes both public and market provided childcare which might shape lone mothers' employment opportunities in very different ways. Nevertheless, in general the empirical analyses show a negative impact of existing gender inequalities on lone mothers' lives increasing their risk of poverty. As a consequence, policies which strengthen gender equality in the labor market and the welfare state would not only help to reduce the gender gap, but particularly to improve the living conditions of lone mothers and their children. This broad cross-sectional analysis wants to be understood as a starting point for future longitudinal research with fewer country cases and a more in-depth institutional analysis.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **Appendix**

**Fig. 8.A1** Coefficients for full time and part time employment on lone mothers' poverty risks from single country regressions (Linear probability models) (Note: CIs displayed at the 95% level)

**Fig. 8.A2** Coefficients for children up to 3 years and up to 6 years on lone mothers' poverty risks from single country regressions (Linear probability models) (Note: CIs displayed at the 95% level)

#### **References**


Christopher, K. (2002). Welfare state regimes and mothers' poverty. *Social Politics, 9*(1), 60–86.


**Sabine Hübgen** is a PhD student at the WZB Berlin Social Science Center. Her research examines explanations for lone mothers' poverty risks focusing on gender inequalities in the welfare state and the labor market, social composition and prevailing gender norms.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Chapter 9 Income Trajectories of Lone Parents After Divorce: A View with Belgian Register Data**

**Dimitri Mortelmans and Christine Defever**

#### **Introduction**

The rise in lone parenthood is one of the most striking symbols of the twenty-first transitions of family life in the Western world. At present, the classic married twoparent family is still regarded as the dominant family norm, although the statistical reality has surpassed the symbolic one. In a few decades, family forms will have such a huge diversity that no dominant, classic or modern family forms will be discerned. We will observe a multitude of family types (Pasteels et al. 2013). The rise of the Western welfare states and the shifts in values we observed during the past four decades started a huge turn in both the composition and the meaning of families. The theory of the Second Demographic Transition (SDT) sees a ideational shift as the driving force behind the changes in demographic behaviour (Lesthaeghe 1995, 2002; Van de Kaa 1994). First, a decline in the birth rates have decreased the size of today's families. Second, the type of relationships in which children are raised has changed. The dominant position of marriage is eroding and the new millennium sees a sharp increase in cohabitation throughout Europe. If people do marry, they marry at older ages (De Wachter 2013) as we observe with parenthood. Postponing parenthood is also often renunciating it leading to smaller families (Bulckens et al. 2007).

The transition this chapter is focusing on, is the rise in divorce and the subsequent increase in the number of lone parent families that often originates from that life course transition. Also this transition finds its origin in the ideational and economic shifts of recent history (Deboosere et al. 2011). Belgium is particularly suited to study lone parenthood since the rise in divorce has quadrupled in the past 30 years and is still increasing today (Corijn 2005; Mortelmans et al. 2011). When comparing

D. Mortelmans (\*) • C. Defever

CLLS, Faculty of Social Sciences, University of Antwerp, Antwerp, Belgium e-mail: dimitri.mortelmans@uantwerpen.be

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_9

divorce figures, Belgium has one of the highest dissolution rates in the European Union. In 50 years, Belgium changed from a Catholic marriage-prone country to the most unstable marriage market (see also: Mortelmans 2013; Mortelmans et al. 2009b). Within the country, regional differences between Flanders and Wallonia are considerable whereby Flanders is slowly closing in on the highest divorce figure in the southern part of the country (Mortelmans et al. 2009a). We study Flanders partly because of the availability of data but also because Flanders has a unique combination of welfare measures and labour market policies that add to the aid of lone mothers to cope with their precarious (income) position. The longitudinal data of the CBSS (see methods section) also allow us to gain reliable insights in the income trajectories of lone parents. Income measures in surveys are always a source of concern and looking at income in small populations like lone parents is an even bigger challenge (Frick and Krell 2010).

In this chapter, we first elaborate on the evolution of lone parenthood in Belgium. We illustrate how this family type rose in importance during the last decades. Next we look at earlier studies that document the socio-economic position of lone parents. A third paragraph introduces the study of this chapter and more specifically the empirical definition(s) of lone parenthood. Before we come to the results and show the income trajectories of lone parents in Flanders, we first need to make clear how administrative data determine our scope. In our analyses, we adopt a life course perspective looking at lone parenthood in a longitudinal fashion: how do lone parents cope with their financial position and who succeeds in tackling the financial downturn after a relationship breakup?

#### **The Rise of Lone Parenthood in Belgium**

The number of lone parents has risen considerably. Several studies have made estimations of the number and evolution of this family type during the past decades. Differences in estimation are often due to differences in definitions or data available. Instead of choosing one single data source, we opt for presenting a multitude of estimations and perspectives in order to get a picture of the lone parenthood prevalence in Belgium and Flanders.

The most important household type in Belgium is the one-person household. The proportion of people living alone rose by 38% between 1990 and 2007 (Lodewijckx 2008). In comparison, the share of married couples with children decreased with 20% and the share of non-married couples with children nearly quadrupled. The reason for the sharp increase in one-person households is the aging population of the country. Most of these single person households are widowed persons. Among the multiple-person households the share of couples (with children) is decreasing while the number of single parent households is sharply increasing (Lodewijckx 2001, 2004a, b). In Flanders, the number of parents raising children on their own increased from 146,800 in 1990 to 205,100 in 2007 (Lodewijckx 2008). When taking figures from the census 2011 (not completely comparable to the study of Lodewijckx), we observe 518,700 lone parent households in Flanders (a share of 9.3% of all households).

When looking at these figures from a child's perspective, we observe an increase from of 117,000 children living with one parent in 1990 to 193,000 children in 2007 (Lodewijckx 2008). Again, the census is more difficult to compare but in 2011 a total of 362,818 children (in Belgium) or a share of 19.4% of children under 15 are living in a single parent household (Swennen and Mortelmans 2015). We do need to take into account that these figures are momentary cross sections of the population. After a relational break-up, people often get new relationships. Not all single parents in the official statistics are without a partner. We estimated that in Belgium, eight out of ten divorcees will repartner, whether or not living together (Pasteels and Mortelmans 2013).

Lone parenthood is strongly gender determined. The vast majority of children living with one parent, are living with their mother. Valgaeren (2008) estimated that one quarter of the lone parent household were lone father households and three quarters were lone mothers. In the census of 2011, we find comparable figures: 21% versus 79%. Despite the huge growth in lone parent families, the gender gap has not shifted. When looking at all mothers, the ONA observes huge differences according to the regions in Belgium. In the capital region of Brussels, nearly one third (28.17%) of all women with children are heading a lone parent family. In Flanders only 13.65% is head of a lone parent family with 22.99% in Wallonia (Office national d'allocations familiales pour travailleurs salariés 2008). For fathers, the total percentage amounts only to 3.44% with again a higher share in Brussels and the lowest in Flanders.

Since the 2007 change in the Belgian divorce legislation, the share of coparenthood is sharply increasing. This is due to a new regulation that obliges judges to first investigate the possibility of co-parenthood. It is difficult to integrate this co-parenting in the figures of the prevalence of lone parenthood. Some indications can be found in the study of the *National Service of Child Support of Workers* (RKW). Between 2007 and 2011, the number of entitlements of child support in case of co-parenthood increased with 30.8%. In 2011, 60.5% of all divorced parents entitled to child support had a co-parenthood agreement (RKW 2011). The Divorce in Flanders study (Mortelmans et al. 2011) estimates the percentage of co-parenthood at 21.4%. Also this study finds an increase over the years, due to a change in legislation. The shared co-parenthood has risen from 6.8% for divorces before 1995 to 27.1% in the period after 2006 (Sodermans et al. 2011).

#### **Income, Poverty and Lone Parenthood in the Life Course**

The sociology of the life course stresses temporal shifts in the lives of people. Often, cross sectional data are the only data available forcing us to one-time snapshots of families. The huge advantage of the registers is that we can observe the financial situation of lone parents over time. It allows us to see which strategies they use to

keep up when the income position becomes difficult after a break-up. This process is a reaction with a temporal change and is termed a coping strategy in the literature. We define a coping strategy as '*a behavioural pattern of an actor to deal with problems, referring to the usual and institutionalized ways of feeling, thinking and acting in such situations*' (Boeije and Nievaard 1995). When looking at the financial coping behaviour after a divorce, we discern two main coping mechanisms: finding a new partner and changing ones labour market behaviour as two ways that positively influence the financial well-being of an ex-spouse (de Regt et al. 2010, 2012; Jansen et al. 2009). We look at these two strategies from the perspective of a lone parent. We assume that also in these cases, a new (co-residing) partner or an increase in labour market participation will help in reducing the (sometimes) difficult reduction in financial position.

The register data in this study does not allow us to relate the gross income of lone parents to poverty lines (calculated on net incomes). But using the EU-Silc survey data, Van Lancker and colleagues were able to estimate the poverty risk of lone parents and compare it with other family types (Defever et al. 2013; Frans et al. 2014). Using the 60% of the median income, the Belgian poverty line of 2010 is at 12,005 Euro while the Flemish one is slightly higher (12,592 for a single person). The mean poverty risk in the country is 15.3% with huge regional differences between Wallonia (19.2%) and Flanders (9.8%). However, when comparing the poverty risk relative to the regional poverty line, the Flemish poverty risk (12.4%) is closer to the one in Wallonia (15.3%).

Van Lanker and colleagues show that the poverty risks in Flanders are highly dependent on the family type somebody is living in. Couples with children have the lowest poverty risk (5%). Single persons have a considerable higher risk (14%) but still lower than lone parents (19%). When looking at the total population of people below the 60% median income, lone parents make up about 10%. Even though the poverty risk among couples with children is lower, in population figures, they are 27% of the total population under the poverty line (Defever et al. 2013; Van Lancker et al. 2012).

One explanation for the increased risk in poverty is the low work intensity. By definition, a lone parent is the only potential breadwinner. In a couple the risk at poverty is different since one partner can be unemployed or absent from the labour market and still rely on the income of the second partner. Lone parents need to work in order to provide income to the household. Earlier studies have shown for Flanders that lone parents are lower educated, more unemployed or absent from the labour market (Cantillon et al. 2004; Mortelmans and Dewilde 2008; Valgaeren 2008). Also in Brussels and Wallonia, a lower work intensity has been observed (Observatoire Bruxellois de l'Emploi 2009; Office national d'allocations familiales pour travailleurs salariés 2008).

An important advantage of using register data is also the possibility to include the gender dimension in the discussion. Often, lone parents are underrepresented in surveys and men within this group are even more rare and invisible. The gender differences are visible both on the labour market and in the household. In general, Belgian women are less active on the labour market. The difference in employment rate between men and women decreases over time but still remains substantial. In 2012, almost 67% of the men (15–64 years) were working, compared to merely 57% of the women. On average, in 2013, a Belgian woman also earned 10% less per hour worked than a man. Due to a large share of part-time working women, the pay gap rises to 23% on an annual basis (Van Hove 2013). The issue of these inequalities becomes apparent when the relationship ends in a divorce or a break-up. Different longitudinal studies have shown that the financial situation of men is relatively stable or even ameliorates. For women, the evolution shows a negative trend in all studies (Andreß et al. 2006). In addition, their situation is also improving in a slow pace, taking a considerable number of years to overcome the financial consequences of divorce (Jansen et al. 2009).

#### **Belgian Divorce Legislation and Financial Consequences of Divorce**

When considering the income position and living conditions of lone parents, it is necessary to briefly show how people end up in this family form. Not that we will present the Belgian divorce legislation in extenso, but particularly the major law reform of 2007 is essential in the rise in lone parent but more specifically in the change in post-divorce material circumstances. The change in 2007 was probably the biggest reform since 1994 (first reform) and 1804 (original law on divorce).

The philosophy of the first divorce law was to protect the basic societal functions of the marriage (procreation, socialization and material welfare). The 2007 reform changed this philosophy by considering the marriage now as a private contract. Not a commercial private contract but a contract originating in love and decomposed when the love is gone. Whenever one of the partners considers the marriage as over, both partners are now given the opportunity to end the marriage with '*no strings attached*'. The law initiated the no-fault divorce and aimed for a quick divorce assuming this would lead to less conflict. Partners are given a 'right' to divorce and no longer need to prove that the other partner has made a mistake in the marriage (adultery for example).

But the divorce procedure was not only shortened, the consequences of divorce have also changed. Before the reform, the so-called innocent husband was entitled to alimony. With the introduction of the no-fault divorce, the rights to alimony also changed into a more universal right to alimony. The only exceptions to the universal right are serious situations like domestic violence whereby a partner can lose the right to alimony. But unlike the universal character of the access to alimony, the new divorce law introduced the time-based alimony payment. Alimony can only be paid for a period that not exceeds the length of the marriage itself. The previous principle of a life-long alimony right, was abandoned. Partners are incited to become active on the labour market and take care of their own income. This has huge implications for lone parents because not only their alimony-income is reduced in time, the new law also changed the definition of the access to the right which implied that the

amounts due are lowered (Cuypers et al. 2008). The child support regulation has not been changed in 2007 but in the Child support objectification law of 2010. Child support can be determined by the judge or through a proposal of the both ex partners. The height of the child support can vary considerably with no standardized calculation methods available (Claessens and Mortelmans 2015).

Other changes that are relevant for this chapter is the change in co-parenthood. With the introduction of this law, a judge is now obliged to examine the possibility of a shared residence (co-parenthood) as the first option for the children's residence. The classic arrangement (only a weekend with one parent every 2 weeks) is still popular but the new law started a huge shift towards co-parenthood. As a consequence, children are more residing with both their mother and their father (often they have a week on-week off arrangement). In these cases, the number of lone parents rises but the amounts of child support among partners decreases since both parents take care of the children.

#### **Aims**

The focus of this chapter lies on the financial trajectory of lone parents after divorce. We aim to bring insights in the financial life course of lone parents by studying their income trajectories over time. The longitudinal character of the register data allows us to do so. Having a large data set, we also explicitly adopt a gender perspective on the finances by analyzing both lone mothers and lone fathers. This will increase our insights in the gender gap among lone parents. Since income trajectories are to a large extent influenced by the household composition, a third aim of this study is to gain insights in the household trajectory of ex-partner after divorce and specifically in the entry and exit of parents in the lone parenthood status.

Based on the theoretical considerations of the previous paragraphs, we come to the following hypotheses:

*H1 Women have a higher chance of entering the lone parenthood state compared to men and they have more difficulties leaving the status.*

*H2 In general, divorced men experience less income loss after divorce compared to women.*

*H3 The income loss of lone parents will be bigger compared to single or repartnered ex-spouses and within one parent, differences will be larger for lone mothers, compared to lone fathers.*

*H4 Staying lone parent will have a more negative impact on the financial life course than making the transition to singlehood or a repartnered status. Again, differences will be larger for lone mothers, compared to lone fathers.*

*H5 For lone mothers, increasing their labour market status will be most beneficial to their post-divorce income trajectory. Unemployment and inactivity have the most detrimental effects on the post-divorce income.*

#### **Data and the Identification of Lone Parents**

In this study, we use the datawarehouse '*Labour market and social protection*'. This datawarehouse is managed by the Crossroads Bank of Social Security (CBSS)1 which clusters register data of social security agencies (see also: Mortelmans and Pasteels 2013). We obtained a random sample of persons living in Flanders and married or cohabiting on December 31, 2003 (T − 1), but no longer living in the same household on December 31, 2004 (T) (they have experienced a relational breakup or a bereavement during 2004)2 (Defever and Mortelmans 2011). The sample contains 25% of all divorced or widowed men and women in the year 2004. For our analyses, we will use only persons that have been divorced and are available to the labour market (18–59 year) in 2004. This entails a sample of 17.044 divorcees. These sample members were followed from 2003 (T − 1) until 2008 (T + 4). All results on sample members in 2005 concern the situation of persons who are divorced at 31 December 2005. Since we only have one demographic position for each year, the status of lone parents is clearly a snapshot in time.

For each household, we calculated the gross yearly income. This household income is composed of all income from labour and social security remittances of all members of the household, but without any taxes or social security payments. The amounts from labour, remittances and pensions are available in the CBSS database. The amounts for minimal income wages and child support were estimated from the data available. Child support was calculated with the number of children in the household, the year, ages of children, household composition and labour market participation of parents and children. Minimal wages were estimated based on the composition of the household of the person entitled to the minimal wage. Since the CBSS does not have any reliable income data (from labour) for self-employed people, no household income was calculated for families in which one of the members was self-employed (this is 19% of the sample of divorcees).

To make the incomes comparable between different household compositions and sizes, we use a modified OECD-scale (Hagenaars et al. 1994). A household income is weighted by the number of adults and the number of children living in the household. The head of the household has a weight of 1. Each additional adult older than 13 years has a weight of 0.5 and each child below 13 years receives 0.3. A lone parent with three children will show a smaller household income in our analyses than a lone parent household with one child if the non-standardized income of both households is identical.

The CBSS data uses the LIPRO typology – developed by the Dutch Interdisciplinary Demographic Institute (NIDI). This typology contain seven individual household types (next to one so-called collective household) (Lodewijckx

<sup>1</sup>We wish to thank the Crossroads Bank of Social Security and more particular Chis Brijs for their cooperation in this research.

<sup>2</sup>For a divorce this means that the person was no longer married or co-residing with the partner on December 31, 2004 with the same partner they lived with on December 31 2003 but they can be living with a new partner.


**Table 9.1** Number and share of lone parents on December 31, 2004, according to the LIPRO and adapted definition, in absolute numbers and percentages

Source: Datawarehouse 'labour market and social protection' – Crossroads bank of Social Security, own calculation

and Deboosere 2008). A single parent in the LIPRO typology is defined as "*a man or woman heading a household and not co-residing with a partner but co-residing with one or more (biological or step-)children*" (Lodewijckx 2004a). This is a very broad definition because no limitations to age, civil status or potential income of the children is included. A single woman living together with her 40 year old divorced son is considered to be a single mother in this definition. Therefore, we define a lone parent in this chapter as a person heading a household alone with no partner and (biological or step)children that are younger than 18 year or younger than 25 when they do not have a labour income. Stated differently, we do not consider a household as a single parent household when: (1) there is a child older than 24 in the household, (2) a child between 18 and 25 having a labour income of its own, (3) a child younger than 18 that is not a biological, nor a step child of the reference person.

This definition is more strict than the definition of the LIPRO typology. Table 9.1 gives an overview of the households we lose from the CBSS when using this definition of lone parents. The percentages are based on the administrative data recorded at December 31, 2004.

According to the LIPRO typology, 25% of all persons experiencing a divorce in 2004 are heading a single parent household after the divorce. If we take our more limited definition, the percentage drops to 20%. One significant reason for the drop is the labour income of children between 18 and 25 year (9% of the lone parents according to the LIPRO definition). The same drop is discernible among the households with children older than 25 or with children that are not related to the single parent (again 9% of the lone parents according to the LIPRO definition). Only a small percentage (3%) could not be unambiguously determined with the data that was available. We decided to consider these cases as non-sample members in our study.

The focus of our analyses are individual persons and not households. Therefore, we will discuss 'lone parents' (or single fathers and mothers) or 'heads of single parent household'. Persons living alone in a household – without partner or children in the household' will be termed 'singles'. If we consider 'repartnered persons', we refer to persons that have a new partner living in the household. The register data do not capture (romantic) relationships until somebody actually moves into the household.

The analytical strategy will follow the hypotheses put forward in paragraph 5. First, we look at the trajectories in household composition 1 year after divorce (T + 1) and again 3 years later (T + 4). In addition, survival curves will allow us to zoom in on the duration of lone parenthood over time. The second analysis concerns the gendered patterns in income after divorce. A first analysis will compare lone parents with other household types. All trajectories are described from the year before the divorce (T − 1) up until 4 years after the break-up (T + 4). The same analysis is repeated for those who start as lone parent in T and we look how differences in household composition at T + 4 influence the income trajectory. The final analysis looks at the position on the labour market. Since the lone parents have too little variation in their labour market trajectories, we only focus on lone mothers in this last analysis. Again, a dynamic comparison is made between T − 1 and T + 4. We look at income effects for women who do not changes their labour market position, and who decrease/increase their attachment to the labour market.

#### **Results**

#### *Partner Trajectories in Lone Parenthood*

In this study, we selected all divorces in 2004. As mentioned before, these persons were living together with their partner in 2003 but that partner has left the household at the end of 2004. In the study, we follow a subset of these divorces. Only those who are lone parent in 2004, are adopted in the empirical sample. One year later (T + 1), we see that already one third of the lone fathers have changed their household position. Either the children (12.8%) are no longer living with their father (presumably they are staying with the mother) or the lone father started living together with a new partner (14.6%). For lone mothers, having no children anymore is rare (1.8%). Women hardly see their children move into the fathers household. On the other hand, having a new partner in the household is much more common. About one fifth of the lone mothers (17.4%) find a new partner within 1 year.

When observing the lone parents 4 years later, we see that less than half of the lone mothers (45.9%) and only one third of the lone fathers (31.4%) remain the single head of their household. Figure 9.1 illustrates this longitudinal pattern by means of survival curves for men and women and the hazard function. The survival function of men is lower than for women, indicating a quicker exit as lone parent than women. The hazard rate shows a decrease in the hazard of leaving the lone parent status until 2 years after the breakup. After 2 years the hazard of entering a new household type are increasing again.

In our register, women do not leave the lone partner status as easily as men. However, they seem both to find a new partner in about one third of the cases (see Table 9.2). The finding is not confirmed with survey data. For Flanders, the Divorce in Flanders study found a lower repartnering chance for women (Pasteels and Mortelmans 2013). Comparing with older research of de Graaf and Kalmijn (2003)

**Fig. 9.1** Survival estimates and hazard function of lone parenthood by gender of the lone parent

the registers observe an acceleration of repartnering patterns. Both men and women repartner in a quicker pace than one or two decades ago. This might be contradictive to the observation that the number of lone parents is increasing but that is not necessary so. Despite the huge repartnering speed among lone parents, the influx of divorcees in the lone parenthood status has a net effect of increasing the total number of lone parent households.


**Table 9.2** Evolution of the size and proportion of household types after divorce among lone parents in 2004, to gender, in absolute numbers and relative %, 2005–2008

Source: CBSS Datawarehouse, own calculations

a The category 'Different' brings together all persons that do not resort under one of the previous four categories. For example somebody who goes living with a friend, an older person who moves to a nursing home, a person who is single parent according to the LIPRO typology but not according to our definition (see above), etcetera

#### *Income Effects of Repartnering*

To compensate the loss of income, lone parents can either repartner or change their labour market participation (if possible). One can increase the number of hours, change employer, etc. Changing one's work can be driven by income motives but can also be an answer to a work-life conflict arising from the break-up.

Research has shown that not repartnering after divorce has the worst negative financial outcomes for women in the long run (Manting and Bouman 2006). Also Jansen et al. (2009) found that, for lone mothers, a new partner has more beneficial financial consequences than changing ones labour market hours. Also, working more hours is not possible in all situations. Geurts (2006) has shown that young single parents (under 30 years) have young children (under the age of 2.5 years3 ). These children cannot (yet) go to school, making the combination between work and care difficult. Only when children are older than 6, the labour market participation of lone mothers equal or surpasses than of mothers with a co-residing partner. Also one's education and the history of labour participation plays an important role in the activation chances after the break-up (for an overview, see Valgaeren 2008).

Tables 9.3 and 9.4 give insight in the evolution of the post-divorce median yearly income trajectories according to partnership status. Table 9.3 gives a view from the start onwards. We follow those who started in T (December 31, 2004) as lone parents, single parents or co-residing parents and we follow these men and women for

<sup>3</sup>From the age of 2.5 onwards, children in Belgium are admitted to the official and free) kindergarten. That means that parents have better opportunities to combine work and care.


**Table 9.3** Evolution of the total gross yearly income of households (corrected for inflation and OECD-HH) for different household types in the year of the divorce (2004 = T), by gender, 2003–2008, in Euro and in % (difference with T − 1, 2003)

Source: CBSS Datawarehouse, own calculations



Source: CBSS Datawarehouse, own calculations four consecutive years (independent from changes in this period). Table 9.4 on the other hand gives an overview of the evolution of income of men and women who started after divorce as a lone parent, single parent or co-residing parent *and* remained in that status for 4 years. From the numbers in the two tables, we can derive that stability in these statuses is rather scarce. Especially the lone parenthood status among men and the single status among both genders is scarce. The N becomes rather small in these cases but nevertheless, the income trajectories show quite interesting evolutions.

We start with Table 9.3 and look at the income loss after the transition (T − 1 to T). Everybody, except single fathers see their income decreasing. The differences however are substantial. Women, whether as lone parent or as single lose 40 or 25% respectively of their pre-divorce income. For lone father, the loss is as limited as mother who immediately repartner after divorce. Off course, these are median income evolutions. At the tails of the distribution losses are more substantial. Repartnering clearly is the most optimal pathway to recover from the initial income loss.

When we look at the evolution, 4 years later, we observe only minor changes for those households in which the divorcee started living with a new partner after the break. Also single fathers have a quite stable income trajectory during these 4 years. Single fathers see an increase in income compared to the pre-divorce situation, especially after 2 years. This is in line with the repartner patterns we have observed earlier. The situation of single women and lone mothers is more problematic. Even though the initial drop in income is compensated over the years, the income loss 4 years after the break still amounts to 10%. Both regain income (by repartnering or increasing their labour market participation) but the loss is still present after 4 years.

When we look at the stable trajectories over 4 years (Table 9.4), we see similar trends: less income loss for men and more loss for lone parents. Even though we need to be careful due to the low numbers, the differences are much more outspoken than when we do not control for stability of the trajectory. When women remain lone parent for 4 years, they can only reduce their income loss to −20.7% instead of −11.5%. For single women, the difference is almost the same: still 17.1% loss after 4 years instead of a loss of 10.9%.

#### *Income Effects of Labour Market Participation*

Finding a new partner clearly makes a difference in the income trajectory of a lone parent. But as de Graaf and Kalmijn (2003) and Pasteels and Mortelmans (2013) showed, repartnering is more difficult in case of lone parents as the children makes them less attractive on the repartner market. In this paragraph, we look at the labour market attachment of lone parents. We limit our analyses to women. The reason behind this decision is straightforward: there is almost no variation in the male labour market trajectories. Nearly all men work in our sample and they all stay at work (full time) after the break.

In Table 9.5, we look at the labour market situation after 4 years for women who became lone parent after their divorce. We compare the situation at time T (2004)

**Table 9.5** Evolution of the total gross yearly income for households (corrected for inflation and OECD-HH) for women who were lone parents in the year of divorce (T, 2004), by labour market evolution (T − 1 compared to T + 4), in Euro and in % (difference with T − 1, 2003)


Source: CBSS Datawarehouse, own calculations

with the position 4 years later (2008). In Table 9.5, we look at stability, increase in participation and decrease in participation at the labour market.

For all lone mothers, we see a decrease of their income after their divorce. In most of the cases, the decrease is around 40% as we mentioned earlier. Only in the case where women worked full time, the drop in income is clearly smaller in year T. When women do not change their activity on the labour market, their initial position keeps determining their income trajectory. As could be expected, a relative lower reduction in income is observed among lone mothers who remain in full time labour. Their median yearly income is also the highest of all women in this sub sample.

To our surprise, part time work did not have the protective effect we expected. In fact, staying out of the labour market (inactive or other) and especially unemployed lone mothers have a much more favourable income trajectory than the part time working women. Even though their income level is well above the unemployed or inactive women, their relative earnings are not restoring as well as the other women. Off course, the precariousness of the unemployed and the inactive women is much bigger as can be seen in their median yearly income. That level lies about one third or more lower than the income level of the part time working women.

When we look at the women who increase their labour market participation (ether or not from zero), the picture is consistent and clear: starting to work or working more decreases the income drop after divorce. In addition, women build up pension rights (Ponnet and Mortelmans 2008) which in the long term assures them a better position (at old age). The reverse is also expected: if the woman loses her job, her income position deteriorates significantly. The unemployment benefit protects women to a great extent. Especially women who are not entitled to an unemployment benefit (and who become inactive without any benefit) suffer severe income loss in a 4 year period and end up with a very low median income.

#### **Conclusion**

When we compare the income position of lone parents in Belgium with those in other countries, we find a relatively favourable situation. The poverty risk is rather limited in comparative perspective and Van Lancker et al. (2012) showed that it is also relatively stable over time. Nevertheless, within the context of the country, lone mothers show a higher than average poverty risk (Defever et al. 2013). Financial consequences of divorce are still highly gendered: women are more often the head of a lone parent household and their income drop is also significantly higher than that of lone fathers.

The poverty figures are only one side of the story. They show the instantaneous picture of who is above or below the poverty threshold. That is important for policy. In this chapter, we looked beyond this cross sectional approach and introduced a life course perspective. Looking at the dynamics behind lone parenthood enriches the view on lone parenthood by not only focusing on poverty and loss. The finding of an increased poverty risk does not help fight against the actual poverty when we do not have insights in the pathways of lone parents. Our point of departure is that people are not frozen targets after divorce but active agents. Even if a relational break-up lead to a severe income loss, people will start coping with the loss. Diminishing consumption levels is one option (out of scope of our study) but repartnering and work are two other coping mechanisms often used. Register data like those from the CBSS are excellently suited to look at dynamics on both the repartner market and the labour market.

A first finding concerns the status of lone parents. Men have the highest odds of leaving the lone parent status quickly (H1). Earlier, we found that men without children have high repartner chances. This study shows that even with children in the household, men repartner more quickly than women. Lone motherhood on the other hand is also not a stable state over the life course. After 4 years, more than half of the women who became lone mother after divorce, have found a partner (or became single because the children left the household).

In line with previous studies (Andreß et al. 2006; Jansen et al. 2009) we also find general post-divorce differences in income favourable for men and negative for women (H2). Within the group of divorcees, it is the group of lone parents and more specifically lone mothers who pay the financial price of the breakup (H3).

Finding a new partner turns out to be a successful strategy for lone mothers in terms of financial well-being (H4). For men, the new partner does not outweigh the single status in terms of finances. Especially because new partners often bring along new children which further increases the costs in the new constellated household. On the other hand, studies have shown that new partners have beneficial effects on the mental and physical health of divorced men (Bracke et al. 2010; Symoens et al. 2013). In the long run, new partners may protect men from an early entry in formal care provisions or emotional loneliness at old age.

The dynamics of the labour market are studied in a different chapter in this volume (see Struffolino and Mortelmans, Chap. 12, in this volume). This chapter focussed on the financial outcomes of changes on the labour market. For men, these dynamics and therefore also the consequences are nearly absent since they predominantly work full time and stay in that status during our observation window. For women, changes in the activity status do lead to better financial outcomes. Increasing ones labour market participation is the most successful strategy but this implies that the circumstances at home allow such an increase or in case of non-employment that the lone mother succeeds in finding an appropriate job (H5). The results on women who decrease their labour market attachment could reveal a double reality. On the one hand, this might be a deliberate strategy to cope with the work-life conflict lone mothers are confronted with. In that case, the strategy is deliberately chosen and the loss of the financial means can be a calculated risk in order to reconcile both life domains. On the other hand, the work-life conflict may also be the cause of the decrease. Women might lose their job or might be forced to reduce their working hours.

Looking at the prevalence and the dynamics of lone parent families, it is clear that this household type needs to stay a concern for policy makers. Even though the

divorce trend is slowly stabilizing, the size of the group of lone parents is considerable and its size will not likely decrease in the coming decade. Earlier longitudinal analyses have shown that the first 6–7 years are difficult for most divorcees (Jansen et al. 2009). There is not only the loss of a partner and (often) one additional income. There is also the cost of the divorce itself: lawyers, moving, tax on buying out a house, new household furniture. A policy that wants to help in this transition needs to be careful for what we have called the "win for life" principle4 (Cuypers et al. 2008). This provoking term asked attention for the simultaneous character of the problematic "shock effect" in the life course that is caused by a divorce. As our results show, many lone parents succeed in coping with the situation (even though they do not live generously). Policies that want to intervene after divorce need to start with the temporality of the break in the life course and try to attenuate the excrescences of the breakup.

We see two possible pathways. First, policies can enable lone parents (and especially lone mothers) to keep or increase their labour market participation. A sufficient supply of child care and a flexible arrangement in terms of opening hours is crucial in the combination of work and care after divorce. Second, financial interventions are possible to keep lone parents out of poverty during the first (and most difficult) years. This might be done through (temporary) increased child allowances. In Scandinavia, we see traces of this kind of policy showing a successful decrease of poverty among lone parent families (Defever et al. 2013).

The financial situation of lone parents shows an example of how a life course policy can be created. Allowances are thereby limited in time and specifically targeted to risks groups with the purpose of overcoming temporary shocks in the life course due to transitions (life a relational break-up). In that respect, we do not need to see these shocks as mere negative transitions. This study was able to show new insights by using large scale register data. Unfortunately, the absence of the educational level in the analyses hampered our analyses. Future studies with registers could replicate this study and document the way to which education mediates the financial trajectories of lone parents.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

<sup>4</sup> "Win for life" refers to a product of the Belgian National Lottery. The first prize is a fixed income of 1000 Euro for the rest of the winner's life. It was used in a context where policy foresees a lifelong payment from the ex-spouse or the government to support the poorest partner after divorce.

#### **References**


Swennen, F., & Mortelmans, D. (2015). *National Report: Belgium*. Brussels: CEFL.


**Dimitri Mortelmans** is Full Professor of Sociology at the Faculty of Political and Social Sciences of the University of Antwerp, Belgium. He teaches qualitative research methods, applied multivariate statistics and advanced topics in family sociology, life-course sociology and demography. He is head of the Centre for Longitudinal and Life Course Studies (CLLS). His research concentrates on family sociology and sociology of labour. He has published on divorce, new constituted families, gendered labour careers and work–life balance. He is also the main author of the Step in Statistics book series of which six volumes have been published (in Dutch). On qualitative methodology, he published the Handbook of Qualitative Research Methods and Qualitative Analysis with Nvivo.

**Christine Defever** holds a Master's degree in Sociology. She works as junior researcher in the Centre for Longitudinal and Life Course Studies (CLLS). She published a Gender Monitor for the Flemish Government and is currently researching lone parenthood and the transition to adulthood.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Chapter 10 The Economic Consequences of Becoming a Lone Mother**

**Susan Harkness**

A growing body of research links the rise in lone parent families to growing income inequality and poverty (Burtless 1999; Chevan and Stokes 2000; Martin 2006). These studies suggest that changing family structures and in particular the rise in lone parent families accounted for up to two-fifths of the rise in US family income inequality (for a review see McLanahan and Percheski 2008). However, as these studies use cross-sectional data they are not able to tell how well-off lone parent families would have been had they not become lone parents: those that become lone parents have been poor even if they were had not had children, or if they had remained living with a partner.

In this chapter we address the question: to what extent is becoming a lone parent associated with worse economic outcomes for women? We do so by tracking women over a long period of time to see how they are influenced by (i) transitions to motherhood and (ii) partnership status. As authors such as Esping-Andersen (2009) have noted this first transition, the transition to parenthood, and its influence on women's employment and earnings appears to have an important impact of lone parent's later outcomes. Yet while a great deal of previous research has examined the effect of motherhood on women's earnings and employment (e.g. Harkness and Waldfogel 2003) far fewer studies have looked at the subsequent influence that these losses have on lone mothers' economic outcomes. The economic consequences of partnership status and, of particular relevance here, partnership dissolution, have been much more widely studied. These studies, with few exceptions, find that women, and in particular women with children, face substantial income losses as a result of partnership breakdown (Page and Stevens 2004; Jenkins 2008; Brewer and Nandi 2014).

While partnership breakdown and divorce are important routes into lone parenthood, a large share of those that experience lone motherhood become lone parents because they give birth to a child without co-residing with a partner. These "birth

S. Harkness (\*)

Institute for Social and Economic Research, University of Essex, Colchester, UK e-mail: S.Harkness@essex.ac.uk

<sup>©</sup> The Author(s) 2018 213

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_10

lone mothers" make up a large share of those who ever experience lone parenthood; data from the UK birth cohort studies show, for example, that among children born in 2000 over 40% of those who experienced lone motherhood before the age of 11 were born to a lone mother (Harkness and Salgado 2018). This route into lone parenthood has been much less widely studied than divorce or separation although a few US studies have looked at the economic consequences of becoming a teenage mother (Geronimus and Korenman 1992) or having children out-of-wedlock (Sigle-Rushton and McLanahan 2002), both groups with a high risk of being lone mothers at birth.

This chapter uses data from the British Household Panel Survey (BHPS) between 1991 and 2008 to assess the economic consequences of becoming a lone mother. It contributes to the existing literature on lone-parent families in two ways. First, we take a novel approach to assessing the influence of lone motherhood on economic outcomes by tracking individuals over time to see how they are affected both by motherhood and by partnership. Second, the consequences of lone motherhood are considered separately for those that were single at the time of their first child's birth, and for those that experience later separations. The economic outcomes we consider are labour market outcomes (employment and earnings); household income (gross, net and equivalised); and household income composition.

The rest of the chapter proceeds as follows. In the following section we review the literature on how motherhood and partnership influence women's economic outcomes. We then describe the UK context. The subsequent sections describe the data and methods, before presenting the results. The final section discusses the implications of our findings and concludes.

#### **Literature Review**

In all countries lone mother families face a disproportionately high risk of poverty (Gornick and Jäntti 2007). Here we review the evidence of the effects of parenthood and partnership on women's economic situation, first looking at the influence that children have on women's employment, earnings and family incomes and second at the impact of partner absence on these outcomes.

Mothers, whether lone or partnered, face large employment and earnings penalties (see, for example, Harkness and Waldfogel 2003; Harkness 2013). Reasons for this include specialization within the household and reduced work effort (Becker 1985); exchanging jobs for those that are more family friendly (Budig and England 2001); reduced labour market attachment and greater constraints on job choices (Manning 2003), as well as direct discrimination against mothers (Correll et al. 2007). These factors act together to reduce mothers' earnings. Reduced earnings may in turn lead to women opting-out of the labour market or reducing their working hours, particularly when their children are young and where the costs of childcare are high (Gornick and Jäntti 2010). Because children have a large effect on women's employment and earnings they also have a substantial influence on families' disposable incomes, although the size of this effect varies widely across countries. The tax and benefit system also compensates for some of the costs of children, although rarely by a sufficient amount to compensate for losses in female earnings (Todd and Sullivan 2002).

While the arrival of children is associated with reduced female employment and earnings and falls in family income, the absence of a partner can have a substantial additional effect. There is some evidence that marriage matters to women's earnings, having a negative effect on pay (Loughran and Zissimopoulos 2009), although there is much little evidence on how separation affects labour market outcomes. Instead most studies have looked at the impact of separation on income. Using panel data to follow individuals over time, the literature on relationship breakdown invariably finds that women see large and persistent falls in their income following separation (Jarvis and Jenkins 1999; Fisher and Low 2012; Brewer and Nandi 2014) although being in employment and re-partnering both offer some protection against falling income (Jenkins 2008).

While the loss of a partner leads to a sharp fall in income, recent years have seen a rapid rise in the number of children who born to lone mothers. In the UK, recent analysis of the Millennium Cohort Survey data shows that 11% of all children born in 2000 were born to a lone mother, and of those that experienced lone parenthood before the age of 11, one-third had done so because they were born to a lone mother (Harkness and Salgado 2018). Yet in spite of the importance of birth as a route of entry into lone parenthood very few quantitative studies have examined how having a child while alone influences women's economic circumstances. One study that does attempt to do so is Page and Stevens (2004). They estimate the "cost" to birth lone parents of not having a partner by seeing how income changes upon repartnering (although they do not look at the cost of having a child). They find the cost of not having had a partner to be substantial, although they are unable to control for selection effects among those who re-partner which may be important given the positive association between re-partnering and employment and income observed in other studies. Another US study of mothers who had given birth out-of-wedlock concluded that even if both parents had stayed together and had both worked fulltime, even under these optimistic conditions, low levels of human capital among the unwed meant that substantial numbers would remain poor (Sigle-Rushton and McLanahan 2002). There are no similar quantitative studies to our knowledge for the UK.

Finally, there is a burgeoning literature on lone mothers which examines how their employment, incomes and poverty rates have been influenced by reforms to the welfare system (Meyer and Rosenbaum 2001; Grogger 2003; Gregg et al. 2009). This literature rarely looks at how the route of entry into lone parenthood influences outcomes, or distinguishes between the influence of motherhood and that of being single for outcomes. Yet there is some reason to expect that this to matter. For example, those that were previously in a relationship may have been more likely to have placed a lower priority on work if they were expecting to be supported by a partner. Those that have children while alone would have fewer expectations of such support and may therefore be more inclined to maintain their labour market position.

Other factors matter too. The fathers of mothers that separate are likely, for example, to get greater support from the absent parent, both in terms of maintenance payments but also in the care of their children, which may make it easier to hold down a job. These differences are rarely explored in studies of lone parents.

#### **The UK Context**

The UK stands out as having one of the highest rates of lone motherhood and one of the lowest rates of lone mother employment in the OECD, contributing towards high rates of child poverty. This makes the UK a particularly interesting country to study as, although rates of lone parenthood are similar to those in the US (which is where much of the empirical evidence on lone parenthood is drawn from), the institutional and welfare context differs substantially.

In 2011 almost one-in-four children in the UK lived in a lone parent family, one of the highest rates in the OECD (OECD 2014). The rise in the lone parent families is a recent phenomenon, as Fig. 10.1 shows, with the share of families with children that were headed by a lone parent trebling between 1971 and 1995, with a large share of this increase driven by the rise in lone mothers that have never married. Since the 1990s the share of lone parent families stabilized, something that is also reflected in our analysis of the BHPS data which shows little change in the share of mothers who are lone parents between 1991 and 2008.

**Fig. 10.1** Share of families with dependent children that are lone parent families by legal marital status (Source: Reproduced from Berrington (2014, p. 5))

**Fig. 10.2** Share of lone and partnered mothers (age 15–64) in paid employment, 2011 (Source: OECD (2014), OECD Family Database, OECD, Paris (www.oecd.org/social/family/database. htm))

High rates of lone motherhood coincide with very low rates of lone parent employment. In 2011 employment rates were among the lowest in the OECD, at around 50% and lone mothers were considerably less likely to work than those who lived with a partner (Fig. 10.2). These differences persisted even after more than a decade of reforms to the British welfare system which had prioritised lone parents' employment and the introduction of a raft of new policies to provide financial incentives to work (through the introduction of tax credits) and activation policies which aimed to support lone mothers find work. While the generosity of out-of-work benefits for families with children may continue to provide a partial explanation of low rates of employment among lone mothers (OECD 2014), there may be other reasons for low employment rates. First, in the UK, as in the US, lone mothers are "negatively selected", often being younger and with lower levels of education than mothers with children (Harkness and Salgado 2018). For these younger, less-educated women motherhood is associated with large reductions in employment and earnings (Harkness 2016). These low overall maternal rates among the less educated, rather than benefit levels, may provide an alternative explanation for low employment rates among lone mothers.

Low maternal employment rates have an important effect on families' income. It is no surprise then, given the large effect that children have on employment rates of mothers that the relative income of families with children in the UK is also among the lowest in the OECD (Fig. 10.3): couples with children have disposable incomes almost one-quarter lower than couples without, while lone mothers average disposable income is just 40% of that of childless couples. Similarly, low employment

WASACH = working-age head, single adult with children WATACH = working-age head, two or more adults with children

**Fig. 10.3** Relative disposable household income of single parents and couples with children relative to working-age couples without children (Source: OECD Family Database 2014, CO2.1, http://www.oecd.org/els/soc/database.htm)

rates and, for those in work, high rates of part-time work and low wages, mean that a very large share of lone parent families are poor. In 2008 over one-third of lone parent families lived in poverty (DWP 2014).

#### **Data Description**

Data from 18 waves of the British Household Panel Survey (1991–2008) is used to track women over time. In each wave the survey reports data from a nationally representative sample of around 10,000 individuals. Households are followed over time, with children joining the main sample once they reach age 16, and new partners of original sample members are also included. The survey is designed to ensure that the sample remains representative of the UK population over time. Later waves also saw the addition of booster samples were for Scotland, Wales and Northern Ireland. The main BHPS data files are supplemented by data from ESRC Data Archive study number SN3909, which provides supplementary estimates of net income and some of its components (such as benefit income) which have been constructed using definitions that match those reported in UK official income distribution statistics, as well as a set of variables that classify individuals by family type, and family economic status (Levy and Jenkins 2012).

The main concern of the paper is to examine how motherhood and subsequent (or concurrent) transitions to lone motherhood influence economic outcomes. Only those who are observed to become a mother to a first child are included in the sample. The sample is restricted to those who are aged under-55, so as to exclude


**Table 10.1** Sample sizes of those becoming first-time mothers and lone mothers

Source: Authors calculations from the British Household Panel Survey

those who may be affected retirement decisions. The final sample, summarized in Table 10.1, includes 1431 individuals who become first time mothers, and 15,786 observations. Of these first time mothers 356 are also observed to become lone mothers (3748 observations) with 187 of these lone mothers being observed entering lone motherhood directly as a result of a first birth and 166 entering lone parenthood as a result of separation. Some women are observed to become lone parents more than once. Where this is the case we assess how their economic trajectories are influenced as a result of the first transition to lone parenthood. Among our sample, as shown in Table 10.1, the average length of time over which birth lone mothers are observed following a first birth is 6 years, and they are observed for 3-years before the birth. For those that separate, we observe them for an average of 10-years after a first birth and for an average of 5-years after separation. Those that remain coupled are observed for an average of 11-years and for 7-years with children.

In our sample, more than half of those observed to become lone mothers enter lone parenthood as a result of a first birth. While this share is high, it is consistent with data reported elsewhere (see Harkness et al. ibid.). In our data we observe children until, on average, the age of 7 and so we can expect this share to be higher. It is also worth noting that lone parenthood is a transitory state, and where lone mothers re-partner the new relationships they form are often unstable. Our calculations using BHPS data, not reported hers, show what while at any point in time the share of lone mothers who enter lone parenthood as a result of a child's birth is around one-if-five, a similar share become lone parents as a result of a breakdown in relationships within step or blended families.

Finally, one concern may be that non-random sample attrition influences our results. Other studies that look at how partnership dissolution influences income using BHPS data find that non-random attrition does not bias their results (Jenkins 2008; Fisher and Low 2012) which provides some reassurance for the results reported here although of course this remains a limitation of this study.

#### **Methods**

One of the primary concerns of any study of lone motherhood is to assess how selection into lone parenthood influences observed economic outcomes. Using panel data to follow women over a long period of time helps overcome this problem as we are able to observe how these women were faring both before and after parenthood and lone parenthood. Studies of divorce and separation take a similar approach, using panel data to follow the same individuals' over time to see how their economic circumstances change as a result of partnership transitions (e.g. Jenkins 2008; Brewer and Nandi 2014; DiPrete 2002; Smock et al. 1999).

The results presented in this study are descriptive and show how employment rates and incomes change for women before and after the birth of a first child and before and after separation. To look at the influence of a first birth on economic outcomes, and to see whether these outcomes are transient or persistent over time, we report outcomes 1-year prior to a first birth, the year of birth, and 1 and 3 years. To look at the association of these outcomes with separation, we similarly look at socio-economic circumstances 1 year prior to separation, the year of separation and 1 and 3 years after. This allows us to assess the separate influence of birth and separation on the economic circumstances of lone mothers.

The socio-economic circumstances we look at are the shares in employment, full-time employment, that are homeowners and who live with their parents as well as hourly wages and income. The measures of income reported are gross, net and equivalised income (using the modified OECD scale). We also report components of income. Parenthood and lone parenthood may influence not only the earnings of women and, where present, their partners, but also their influence benefit receipt. We report descriptive data to show how women's earnings, (where present) spouse's earnings, transfer income and benefits change around transitions to parenthood and lone parenthood. All monetary values are deflated to January 2010 prices. We trim household income data at 1% to avoid including negative or zero incomes, and to avoid problems of top coding. Finally, to adjust income for needs we use the modified OECD equivalence scale which give the first adult a weight of 1, each person age 14 or over a weight of 0.5 and each child under 14 a weight of 0.3.

Throughout, lone mothers who separate and those who become lone mothers as a result of a first birth are considered separately, and their incomes are compared to those of families where children continue to live with both biological parents. This allows us to get a sense of the importance of prior economic circumstances for the observed poor economic circumstances of lone mothers.

#### **Lone Mothers in the UK: Characteristics, Employment and Income**

The US literature has described the "diverging destinies" of children being brought up in lone parent families and those living with both biological parents (McLanahan 2004). Over recent decades, she argues, the resources (in terms of parental inputs and income) available to those growing up in these different family types have diverged with lone parenthood becoming increasingly concentrated among the less educated. Table 10.2 shows a similar picture for the UK, lone mothers being on average younger at the time of first birth and holding fewer qualifications than mothers who live with a partner. Most studies of lone parents look only at


**Table 10.2** Individual and family characteristics of mothers with partners and lone mothers by birth and separation (full sample and those observed prior to having a first child

Source: Authors calculations from the British Household Panel Survey

characteristics of lone mothers observed at a particular point in time and do not distinguish lone mothers by their route of entry to lone parenthood. Looking only at those who become lone mothers over the period of study, reported in Table 10.2, we see large differences between lone mothers that separate and those that become lone mothers by birth: while those that separate have lower levels of education, are on average younger at the time of becoming a first time parent and are more likely to be home owners or live with their parents than those in couples. The differences between those who separate and who become lone mothers by birth are larger again. For example, one-third of birth lone mothers were homeowners compared to 51% of those that separated and over 80% of those in couples. One-quarter lived with their own parents compared to just 6% of separating lone mothers and 1% of couples. The average age at the time of a first birth among birth lone mothers is 24, compared to 27 for those who separate and 29 years among those who remain partnered. However, there is a large degree of heterogeneity among birth lone mothers in particular. While birth lone mothers are particularly likely to be young and hold no qualifications, there are also a fairly large number of women who are older at the time of a first birth (one-in-five birth lone parents had their first child after the age of 35) and 10% had degrees. Comparing all lone and partnered mothers to the subsamples that are observed before the birth of their first child in Table 10.2 shows strong similarities between the full and restricted samples.

As in the US then, lone mothers in the UK have lower levels of human capital than mothers that remain in couples. This is reflected in their labour market outcomes and incomes, reported in Table 10.3. Average employment rates of lone mothers


**Table 10.3** Employment, wages and income of mothers with partners and lone mothers (by birth and separation), full sample and those observed prior to having a first child

Source: Authors calculations from the British Household Panel Survey

**Fig. 10.4** Kernel density distribution of equivalised net household income, lone mothers and couples with children (Source: BHPS 1991–2008. Income is reported at Jan 2010 prices and is equivalised using the modified OECD scale)

stand at around 50%, compared with 65% among mothers in couples, while weekly earnings (which exclude those with no earnings) are almost 30% lower. These aggregate figure again disguises large discrepancies between those who were birth lone mothers and those that separated after having a first child. In particular, separated lone mothers have employment rates similar to partnered mothers, at 64%. In contrast only 38% of birth lone mothers were employed. Earned income is also much lower for birth lone mothers; so while those that separated have weekly wages around 10% lower than partnered mothers, at £185 per week compared to £205, the earnings of birth lone mothers were only just over half the earnings of those that separated at £106. Yet, in spite of these differences in the labour market outcomes of birth and separating lone mothers, income differences were much smaller. Average needs adjusted, or equivalised, incomes for mothers in couples were £319 a week compared to £226 in separated lone mother families and £203 in birth lone mother families. The distribution of equivalised income in lone mother families, and mothers in couples is shown in Fig. 10.4. Even after adjusting for family size lone mothers' incomes fall far below those of mothers in couples, and their incomes are strongly clustered at the bottom of the income distribution: only a very small number of lone mothers have incomes higher than the median income of couples with children.

#### **Trajectories Over Time: Changes in Economic Outcomes**

The labour market outcomes and income data reported in Table 10.3 show lone mothers do badly. These are at least in part likely to be due to the relatively poor characteristics of these mothers. However, they may also be a result of having become a lone mother, whether by birth or separation. To examine the extent to which this is the case we look at how well these women were faring before becoming mothers and lone mothers.

Table 10.4 shows how labour market outcomes changed around two critical points (i) becoming a first time mother, and (ii) following the dissolution of a partnership. Rates are reported for any employment and being in full-time work (30+ hours) and for three groups; mothers who remain with their partner after birth; those that separate and those who were single at the first birth. The first thing to stand out from this table is the large difference in employment rates that existed prior to becoming a mother. Only 63% of birth lone mothers were working prior to a first birth compared to 76% of those who later separated and 87% of those who remained partnered. Differences between those that become lone mothers and those that do not therefore appear to account for large differences in employment outcomes. Second, employment drops much more sharply for those that go on to become lone mothers in the years just after a first birth: while for partnered women employment rates were 18 percentage points lower after a first birth than they had been the year before, the drop was 29 percentage points for both separating and birth lone mothers. For both these groups employment rates recovered 3-years after the birth.

We find separation to have very little association with employment rates. In the year preceding separation 57% of women were in work, and this fell by less than 1% in the year of separation. Employment rates then began to recover and three years after separation were 9 percentage points higher than they had been before


**Table 10.4** Employment before and after a first-birth and before and after lone parenthood (£ per week)

Source: Authors calculations from the British Household Panel Survey


**Table 10.5** Share of women that are homeowners, living with their parents and who re-partner before and after a first birth and before and after lone parenthood

Source: Authors calculations from the British Household Panel Survey

the split. For full-time employment, we observe a sharp drop in employment following a first birth with only one-quarter of women who remained living with the child's father in full-time work three years later and with the share among those that went on to become lone mothers far lower. Separation had little influence on fulltime employment rates.

Of course, becoming a mother and a lone mother may also be associated with other changes in living arrangements. Table 10.5 shows how living arrangements changed before and after having children and becoming a lone parent. In particular, we look at changes in homeownership, co-residence with own parents; and repartnering. Those that become lone mothers are much less likely to be in owneroccupied accommodation in the year before becoming a first time parent. Among those that do not separate, around 80% are home owners the year before a first birth and this share rises slowly with time. Only a very small share live with their own parents at the time of their first child's birth (6% the year before a first birth and less than 1% three years after). Among those families that going on to separate, just under 3-in-5 are home owners at the time of the first child's birth. These women are also likely to be living with their own parents in the year preceding a birth (1-in-5 mothers), and at the point of separation almost 1-in-10 live with their parents. Those that have children while alone are least likely to be in owner occupied accommodation and are most likely to live with their own parents – around half live with their parents at the time of birth and while this declines over time the share co-residing with their parents remains high, at 1-in-6 three years after a first birth. Finally, around 1-in-10 lone mothers re-partner within a year of birth (for birth lone mothers) or separation, while around one-quarter have re-partnered within three years.

**Fig. 10.5** Income before and after children and before and after separation. (**a**) Remain in couples. (**b**) Birth lone mothers. (**c**) Separating mothers: change following motherhood. (**d**) Changes following separation (Notes: as Table 10.4)

Figure 10.5 looks at how transitions to motherhood and lone parenthood affect income. It illustrates how the income distribution shifts before and after having a first child, and before and after separation. Panel (a) shows how income changes before and after childbirth for those who remain in couples. For this group the income distribution shifts to the left following a first birth with a further fall in income three years' post-birth. Birth lone mothers (Fig. 10.5, panel (b)) have much lower levels of income before a first birth, and while we see a similar leftward shift in the distribution of income following a birth this shift is less marked than that for those in couples. Those that separate have more dispersed incomes than either of the other two groups prior to a first birth. A first birth again leads to a left-ward shift in the income distribution, with the post first-birth distribution resembling that for those mothers who remain in couples. The association with subsequent separations is shown in panel (d). Most notable here is that while income declines following a split across the distribution, the influence on income of separation is smaller than that of having a first child, particularly when we observe income 3 years after the split.

Table 10.6 shows which income components drive the overall change in income. For partnered families, mothers' labour income drops substantially following a first birth and continues to decline over the following three years, partly as a result of


**Table 10.6** Income levels and composition before and after a first-birth and before and after lone parenthood (£ per week) 227


**Table 10.6** (continued) changes in employment, partly because of changes in working hours (in particular the switch to part-time work), and partly as a result of reduced hourly wages. On average mothers' labour market earnings fall by around one-third three years after a first birth. However, within couples, losses in women's earnings around childbirth are to some extent offset by increases in their partners' earnings. In row (2) of Table 10.6 (i) the sum of both partners' earnings is reported. The combined earnings of couples, while showing a small dip around the years of childbirth, recover to a level very similar to their pre-birth levels within three years. In row 3 transfers are added to earned income (row 2). As couples with children receive low levels of transfer income this does not lead to a substantial change. Row 4 adds other income, which includes income from savings and pensions and other household members' income. Income from these sources falls too around the time of a first child's birth, further squeezing incomes. Compensating for these losses however are transfers through the benefit system (row 5). Household benefits, including child related benefits, offset the losses in private income which result from having children. These increases in benefit receipt, together with increased male earnings, mean that the overall influence on gross income of having a first child is roughly neutral. The change in net income is similarly neutral (column 6). Finally, after adjusting for needs, in column 7, equivalised income shows a fall of up to 25% in the years following a first birth with no recovery over the next three years that we observe them for. Among mothers in couples therefore we see sharp falls in own earnings following a first birth but that these losses are to some extent offset by increases in the earnings of their partners and by increases in benefit income.

Next we investigate how the experience of those that become lone mothers differs to those who remain in couples. In panel (ii) of Table 10.6 income changes are shown for birth lone mothers. Among this group, average earnings are much lower before childbirth than the mothers who remain in couples, at £163 per week compared to £309 for partnered mothers. And for these women earnings also fall, by around one-third three years after a first birth, following childbirth. These fall leads to an overall drop in family earnings (row 2), although this loss is slightly lower than that reported in row 1 because some birth lone mothers re-partner. Transfer income (row 3), including maintenance, is also low among this group and does little to compensate for losses in earned income. Other household income, and in particular other household members' earnings, are a particularly important income source for birth lone mother in the years before and after a first birth (row 4), as is household benefit income (row 5). However, over time as birth mothers establish their own homes there are sharp drops in other family members' income. These losses, together with falls in their own labour income, lead to substantial falls in gross income. To some extent these losses in gross income are compensated for through the tax system (row 6), and changes in family structure as women set up their own households reduces income needs. As a result, overall needs adjusted, or equivalised, income falls by around 20% on becoming a birth lone mother and remains at this level in the subsequent year.

The final panel of Table 10.6 (iii) show income changes for our second group of lone mothers; those that become lone mothers by separation. For this group earnings,

while higher than those among birth lone mothers, are lower than for those in couples that remain together and fall by a similar magnitude following a first birth. Following separation earning begin to increase slowly, contrasting with the experience of those that remain in couples whose own earnings decline with time. In row 2 the total earnings of mothers and their partner are reported and again this shows a decline in the couples' earnings following a birth, but that at the point of separation couples' income was higher than it had been prior to the first birth. Separation is associated with substantial losses in earned income because of the fall in spouses' earned income. Transfers, including maintenance payments, play only a very minor role in offsetting these losses. Row 4 also shows a decline in income resulting from other family income which may reflect changes in living arrangements with time. Finally, benefit income has an important impact on reducing the income losses associated with both motherhood and lone parenthood. As a result, for those that separate, motherhood leads to an overall fall in gross household income of around 10%. This recovers to its earlier levels by the year before separation, largely because of gains in partners' income and in benefit receipt. Separation leads to large further losses in gross income with gross income halving in the years before and after separation largely because of loss partners' earnings although there is some additional compensation through the benefit system and increased own earnings reduce this gap with time. Net income also falls around separation (row 6). However, adjusting for needs in row 7 we find that having a first child reduced income by almost 20%, although there is some recovery before separation. At the point of separation equivalised income falls again, by a magnitude only slightly larger than that following the arrival of having a first child.

Finally, comparing the incomes of those that enter lone parenthood as a result of a first birth while single and those that become lone parents through separation, we find that incomes are almost identical for these two groups once they become lone parents. This is in spite of the fact that those that separate have higher incomes before having children, and that upon become lone mothers have higher own labour income and receive more income from transfer payments.

#### **Discussion and Conclusion**

Lone parenthood in the UK is common and strongly associated with low income. The analysis that we present here suggests that differences in lone mothers and those who are partnered partly result from differences in selection. As the descriptive statistics showed, lone mothers, whether they enter lone motherhood as result of separation or through the birth of a first child while alone, are on average younger at the time of a first birth and less educated than those who are never observed as lone mothers. Levels of human capital among birth lone mothers are particularly low. Even before the birth of a first child, rates of employment and earnings of both groups are far lower than those of mothers never observed to be lone parents.

In addition, partnership and parenthood matter. Becoming a first time mother imposes a substantial economic cost on women and families, reducing their labour force participation, hours of work (with many women switching to part-time employment), and wages. The association of becoming a first time mother with the employment rates and earnings of those that also become lone mothers are of a similar magnitude to those for women in couples that remain together. Our analysis therefore questions the idea that low employment rates among lone mothers in the UK are driven by high levels of benefit receipt (OECD 2014). Instead, we find no indication of differences in the labour market behaviour of those that become lone mothers and those that do not: lone mothers respond in a similar fashion to other mothers following the birth of a first child. The low employment rates among lone mothers instead appear to reflect first "selection" into lone parenthood – those that become lone mothers had low employment rates even before they became parents; and second the large negative influence that parenthood has on maternal employment among all mothers in the UK.

Parenthood also has a substantial influence on income. Even among those that remain in couples we see falls in equivalised income, of up to 25% immediately following a first birth. This loss is largely driven by reduced maternal income and increased family size, although this is compensated for by increases in the earnings of male partners and increases in benefit receipt (including child benefit). For those that become birth lone mothers, mothers' earnings similarly fall. For this group however there is no male partner to compensate for the loss in mothers' earnings by increasing their labour supply. Nor do this group receive significant income in the form of transfer payments, including maintenance. This group are instead particularly likely to live with other family members, in particular their own parents, and the earnings of these other household members make up a very important source of income, alongside benefit receipt, in the years before and after a first birth.

Mothers that separate, while having lower levels of human capital, employment and earnings than those that remain in couples, start off with higher levels of income and earnings than those that become lone parents by birth. They are also much more likely to be living independently of their parents at the time they have a first child. On becoming mothers this group too face large falls in their earnings, although becoming a lone parent is associated with slight increases in mothers' earnings. For these women transfer income, including maintenance payments, and benefits are important sources of income following separation. However, upon becoming a lone mother the economic situation of these women converges to become very similar to that of birth lone parents (Fig. 10.5) in spite of their earlier economic advantage and higher levels of earned income and maintenance payments.

The results reported in this study are descriptive and, while they follow women over time, they do not directly account for the effect of differences in characteristics on observed outcomes. In addition, they do not take account of unobservable differences between lone mothers and those with partners that influence "selection" into lone parent or employment. Nonetheless, in spite of these limitations, they indicate that the poor economic circumstances of lone mothers are likely to be only partially driven by the transition to lone parenthood.

Overall we have shown that for all families there is a high economic cost attached to motherhood in the UK. The results reported here echo those of Todd and Sullivan (2002) who find families with children to have significantly lower disposable incomes than those without in the UK. The difference in incomes of families with children and those without is, they argue, largely a result of the effect that children have on women's employment. They show that in countries where mothers remain in work after having children families fare better economically: in countries including the US and Scandinavian countries incomes are actually higher among families with children. Other studies similarly show that high rates of female employment are associated with reduced levels of income inequality (Harkness 2013).

Esping-Andersen (2009) argues that in countries where rates of employment are high among mothers' rates of family poverty tend to be low, particularly because the rate of lone parent poverty falls. For the UK we have shown that falling female earnings following a first birth accounts for around half of the income loss associated with becoming a lone mother among those that separate, with subsequent partnership dissolutions lead to further falls in income of a similar magnitude. Maintaining the earnings position of women at the same level that they had prior to having children would considerably improve the income of all lone mothers. Other studies suggest that this matters not only in the short term; as Brewer et al. (2012) note "family income during the main child-bearing years is an especially good predictor of lifetime income as the consequence of permanent differences in productivity and marriage prospects can be more visible at a time when working is particularly costly" (p. 36).

In the UK many women, and in particular those with lower levels of education, find it hard to sustain employment, and in particular full-time employment, following the birth of a first child. For these women the earnings of a partner are a critical source of income. Yet it is these women, who have the weakest attachment to the labour market, that are also at greatest risk of becoming lone mothers (Gregg et al. 2015). This combined risk of non-employment and lone parenthood means that for many mothers the risk of poverty and low income are likely to continue to remain very high.

**Acknowledgements** We would like to thank the Economic and Social Research Council for supporting this research under grant number ES/K003984/1.

This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **References**

Becker, G. (1985). Human capital, effort, and the sexual division of labor. *Journal of Labor Economics, 3*(1, Part 2), 33–58.

Berrington, A. (2014). *The changing demography of lone parenthood in the UK* (ESRC Centre for Population Change Working Paper Number 48).


**Susan Harkness** is Professor of Public Policy and Director of the Understanding Society Policy Unit at the Institute for Social and Economic Research, University of Essex. Her research focuses on women's employment and earnings, and their influence on poverty and inequality. Her recent research has looked at lone parent families' employment and income, and the influence of lone parenthood on children's outcomes

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Part IV Labour Market Behavior of Lone Parents**

## **Chapter 11 The Role of Informal Childcare in Mothers' Experiences of Care and Employment: A Qualitative Lifecourse Analysis**

**Michelle Brady**

#### **Introduction**

This chapter argues that single mothers who use mixed childcare (concurrent use of formal and informal childcare) have much greater success in entering and sustaining employment compared to single mothers who have a formal-only childcare package (i.e. rely entirely on formal childcare), or an informal-only package. To date the literature on single parents and employment has largely ignored the widespread use of mixed childcare amongst families. While this omission is important for all family types, it is a particular problem for the literature on single mothers because my prior research has found that single mothers who use mixed childcare on average working significantly longer hours compared to mothers who use other childcare packages, after controlling for a range of demographic factors and hours of childcare used (Brady and Perales 2016).1 Further, descriptive statistics show that Australian single mothers who use mixed childcare packages have higher employment rates compared to mothers who use other childcare packages (Baxter et al. 2007).

Drawing on a lifecourse and *caringscapes* conceptual framework and applying them to a qualitative longitudinal (QL) study with 30 Australian single mothers in receipt of single rate Parenting Payment (PPS) (an income support payment) and with at least one child aged less than 7 years, the chapter argues that the observed correlation between hours of employment and childcare package exists because, for most single parents, the childcare offered by formal providers is insufficiently

<sup>1</sup>Partnered mothers who use mixed childcare also work significantly longer hours compared to mothers who use only formal care or only informal care. However, for "for partnered mothers mixed childcare increases employment hours by maximizing the hours of childcare available to them" whereas this is not the case for single mothers (Brady and Perales 2016, p. 1).

M. Brady (\*) School of Social Science, University of Queensland, St Lucia, QLD, Australia

School of Public Administration, University of Victoria, Victoria, BC, Canada e-mail: m.brady2@uq.edu.au

flexible on a day- to-day basis and does not adequately respond to changes that occur across the children's and mothers' lifecourses. Furthermore, informal-only care is unreliable over the longer term. Mixed childcare packages typically provide mothers with access to the longer hours of care usually associated with formal care, but also access to the flexibility usually associated with informal care. This flexibility afforded by the informal care allows these single mothers to adapt to changes across the lifecourse.

The degree to which the flexibility of informal carers is necessary is dependent upon the childcare and employment policy context. In Australia formal childcare is inflexible in three distinct ways. Firstly, formal childcare is generally only offered during standard working hours and very little formal care is offered in the evenings or on weekends. These restricted childcare hours cause problems for the many single mothers who work non-standard hours. Secondly, formal childcare requires families to choose fixed days of care and generally does not allow families to change their days of care from week to week. This requirement to choose fixed days of care generates problems for employed mothers who have rotating shifts. Finally, formal care is inflexible in terms of offering sharply age segmented care. For example, some providers only offer care for children under 2.5 years whilst others do not offer a kindergarten program for 4 year olds. The result is that parents commonly need to change providers multiple times in the first decade of a child's life, and once their child starts high school they usually find that there are no formal care options available.

It is in this childcare context that informal carers (in the context of a mixed childcare package) play a crucial role in assisting single parents to enter employment and to sustain it. Informal carers supplement formal care by smoothening transitions associated with changing childcare needs, assisting mothers with childcare crises, and stepping in to assist with some of the juggling associated with managing work and care work on an on-going day-to-day basis (Brady 2016). Thus, in a mixed care package informal care supplements rather than replaces formal care. In the conclusion of this chapter I argue that these findings suggest that there is a need for the formal childcare system to be more flexible and supportive of mothers' and children's changing childcare needs.

The structure of this chapter is as follows: (1) the current literature on single parents' lifecourses, childcare and employment, and on QL methodologies is reviewed; (2) the Australian work and family policy context is described; (3) the study, including the data and analysis, is described; and (4) I illustrate how single mothers' use of mixed childcare packages shapes their entry into employment/ study, their ability to balance paid employment/study and caring commitments once they enter paid employment/study, and their ability to manage shifting childcare needs across the lifecourse.

#### **Single Parents' Lifecourses and Childcare and Employment**

Existing research finds that single mothers have diverse employment trajectories with some experiencing long periods without employment, others cycling in and out of employment (and on and off income support), and yet others maintaining fulltime employment with few breaks (Barrett and Cobb-Clark 2001; Barrett 2002; Zagel 2013). A single mother's ability to enter employment and maintain it is dependent on her ability to find appropriate alternative carers during the periods she is working (Brady 2010; Cabrera et al. 2013; Scott et al. 2005). Making care arrangements is not a one-off activity but instead once single mothers gain employment they must assimilate this activity with "caring for their children on an ongoing basis, sustained over time" (Millar and Ridge 2009, p. 104). In the United States researchers find that when single mothers who are financially dependent on Temporary Assistance to Needy Families (TANF) move into paid work, they are commonly forced to depend on unstable patchworks of care. This reliance on informal care is due to financial constraints (including lack of access to subsidies for formal childcare), and the low supply of formal care in low income areas (Scott et al. 2005). The unreliability of their care is due to the unpredictability of informal carers over time (Fuller et al. 1996; Scott et al. 2005) with family not always following through on commitments (Fuller et al. 1996). Research also concludes that these childcare difficulties are the key factors that causes single mothers to return to TANF (Anderson et al. 2004).

Due to this unreliability of informal childcare, and its association with poorer child outcomes (particularly for children from disadvantaged backgrounds (Del Boca et al. 2014)); scholars in the UK, US, Australia and elsewhere have advocated for the provision of affordable, high quality formal childcare for single mothers who are moving from "welfare to work" (Brady and Cook 2015). At the same time these scholars recognise the importance of social support in enabling single mothers to sustain employment (Mason 2003; Ridge and Millar 2011; Scott et al. 2005). In the UK Millar and Ridge (2009) argue that social relationships across work, home, school and care settings "may be a key element in employment sustainability" (p. 104), while a review of the literature on single mothers and social support by Cook (2012) argues that families and friends may provide crucial free transport, food, and childcare that help mothers combine work and care.

As I elaborate in more detail below what is missing from this literature is a consideration of the experiences of single mothers who use formal and informal childcare concurrently (i.e. mixed childcare packages). As I have argued elsewhere (Brady and Perales 2016), this is problematic insofar as single mothers who use mixed childcare on average work significantly longer hours compared to mothers who use other childcare packages (after controlling for demographic factors and hours of childcare used). Although McKie et al. (2002) have made a compelling case for examining the temporal dimension of balancing work and care, to date the dynamics of single mothers' experiences of balancing work and care have been underexplored. Elsewhere I have focused on conceptualizing the key roles played by informal carers over time (Brady 2016). In contrast this chapter draws on lifecourse concepts to systematically examine how single mothers' use of mixed childcare packages shapes their employment trajectories over time.

#### **Qualitative Longitudinal Research: Fateful Moments, Timespaces and Caringscapes**

QL research is increasingly being used to understand multiple dimensions of key lifecourse transitions including "welfare to work", or "school to work" (Holland and Thomson 2009; Lewis 2007; Thomson et al. 2002). As addressed in other chapters, lifecourse studies focus primarily on the timing, duration, and cessation (or desistance); as well as order, sequencing, and occurrence of life transitions, life events and trajectories and the relationships (or interlock) between earlier and subsequent life course transitions and trajectories (Macmillan and Copher 2005). Within the lifecourse literature trajectories are defined as "life course dynamics that take place over an extended period of time" and as starting and ending with a transition (Macmillan and Copher 2005, p. 5; Mayer and Settersten 1997). Transitions are embedded within trajectories and are concerned with how people move from one "role to another, begin or cease a course of activity, experience a particular state, or stop doing so"(Macmillan and Copher 2005, p. 5). Furthermore, in contrast to an 'event', transitions are "conceptualized as a more gradual change" (Mayer and Settersten 1997, p. 252).

In the area of work, family and social policy lifecourse research has been overwhelmingly quantitative but the number of QL studies has grown significantly over the last decade (Holland and Thomson 2009; Millar and Ridge 2009; Thomson 2007). QL research makes an important contribution to the lifecourse literature because it enables a focus on dimensions and experiences of time and change that are not possible in quantitative research. Although QL research has in common with quantitative approaches a focus on transitions, it diverges from quantitative approaches in that it seeks to illuminate the perceptions of the individuals involved (Millar and Ridge 2009; Thomson et al. 2002). A particular focus has developed around critical moments or fateful moments, "times when events come together in such a way that an individual stands, as it were, at a crossroads in his existence" these include things such as the "deciding to opt for a particular apprenticeship or course of study" or "giving up one job in favour of another" (Giddens 1991, pp. 112–3; Holland and Thomson 2009; Thomson et al. 2002).

QL research also has the potential to add to the lifecourse literature through a focus on the multiplicity of time, though to date this potential has largely not been exploited. For instance, McKie et al.'s (2002) work has drawn attention to the spatial and temporal dimensions of combining paid work and care. Their work builds on Adam's (2000) notion of *timescapes*, and its associated argument that we should supplement quantitative time (calendars and clocks) with recognition of context (time and space), and the multidimensionality of time (different functions of time). Building on this notion of timescapes McKie et al. (2002) argue that there is a need to pay greater attention to the experience of combining paid work and care across the lifecourse. Within their discussion of caringscapes – "imagining the spatial-temporal nature of everyday life" (p. 918) – they argue we need to pay greater attention to the temporal dimensions of the experience of combining caring and working including:

planning, worrying, anticipating, speculating prioritizing, assessing the quality of care, accessing care, controlling care, paying for care, shifting patterns of work, job (in) security, the potential for promotion, moving home, managing family sources, supporting school work, being involved in the school or care group and so on (p. 915).

In summary then, current QL research extends the quantitative lifecourse literature's focus on the timing or sequencing of life transitions, events and trajectories to examine subjective experiences of transitions, and the specific time and context. Within the research area of work, family and social policy the QL approach also has the potential to highlight how socially embedded individuals experience changing work and care arrangements, needs and tensions.

#### **Australian Context: Single Parenthood, Income Support and Childcare**

Single parents' experiences of entering employment and combining work and care on an on-going basis is strongly shaped by the national context. Some important factors that shape this experience include the prevalence of single parent families, the specifics of childcare/early education systems, the nature of policies around income support for single parents, employees' entitlements, and cultural attitudes to early childhood care (Mamolo et al. 2011; Marchbank 2004).

In June 2012, 22% of all families with dependents in Australia were one parent families (ABS 2013). One parent families with dependents are defined in official statistics as a "person who is not in a couple relationship with anyone usually resident in the same household", but has at least one child usually resident in the household who is aged under 15 years or a dependent student aged 15–24 years who is "attending school or studying full-time at a tertiary education institution" (ABS 2013). A little over half (52%) of single mother families with dependants (0–14) were employed and employment rates were much lower for those with younger children (ABS 2013).

Australia is usually classified as a liberal welfare regime, but it has also been classed as a wage earners' (or antipodean) welfare regime (Castles and Mitchell 1990; Castles 1994; Mitchell 1998). Wage earners' welfare regimes have in common with liberal welfare regimes a focus on targeting income support at those with low incomes and assets and a preference for largely leaving childcare and elder care to families, charities, the market, and compulsory schooling (Esping-Andersen 2012; Kilkey 2000). However, unlike liberal welfare regimes, they also promote worker's well-being through relatively strong regulation of working conditions and wages. Casual and part-time workers in Australia therefore benefit from somewhat better wages and conditions compared to workers employed under these contracts in other jurisdictions.

Overall, Australia's policies around income support for single parents, and early childhood education and care (ECEC), are largely consistent with the tendencies associated with liberal and antipodean welfare regimes. Since 2002 the Australian government has applied compulsory activity requirements to claimants of PPS who had a school aged child. New policies implemented over the last decade have progressively increased the intensity of the activity requirements that are applied to this group. In the week prior to the commencement of data collection for this study the government announced new activity requirements. Australian single parents with a child aged less than 8 years would be allowed to continue to claim PPS and would not be obliged to engage in paid work or education/training. However, those with a child aged over 7 years would be moved to an unemployment payment (Newstart allowance), which is paid at a lower rate than PPS, and they would be obliged to seek at least 15 h of paid work (per week). However, those who were already claiming PPS when the changes were announced were exempted from the requirement to move to Newstart. Thus some of my interviewees were subject to the new activity requirements but none were moved to Newstart payment during the course of my study.

Consistent with its classification as a liberal or antipodean welfare regime, nonparental care for children under school age2 is most commonly provided by relatives (usually grandparents), for-profit childcare centres, and to a lesser degree not-forprofit organizations and family day-care (child-minders) (Australian Bureau of Statistics 2012; Baxter et al. 2007). A little over half (55%) of all children aged 0–4 attend childcare, and of these 40% only use formal childcare, a quarter (26%) only use informal childcare, and a third (32%) attend both formal and informal care (Australian Bureau of Statistics 2012). Patterns of childcare use amongst employed single mothers with young children largely reflect these broader patterns with a third (33%) using formal only, a little over a quarter (28%) using informal only, and a little over a third (37%) using both formal and informal care (mixed childcare) (Brady and Perales 2016).3 These patterns of childcare use are also very similar to those observed in other liberal welfare regimes including the US and UK (Emlen 2010; Hansen et al. 2006; Rutter and Evans 2011).

While the broad patterns of childcare use in Australia are similar to other liberal welfare regimes its approach to ECEC deviates from a typical liberal regime in some significant respects. One importance deviation is Australia's emphasis on tightly regulating the quality of childcare via the Australian Federal Government's National Quality Framework (Brennan and Adamson 2014; Mahon et al. 2012). A second important difference is that unlike the UK, Canada, and the United States; Australia has a national, income-tested childcare subsidy scheme (Childcare Benefit) and a non-income tested childcare rebate (Brennan and Adamson 2014). For the vast majority (95%) of families, government subsidies cover at least 50% of

<sup>2</sup>The age at which compulsory full-time schooling commences varies from state to state, ranging from 4 years and 7 months, to 5 years.

<sup>3</sup>Formal childcare is defined as care purchased via the market and includes family-based day-care, centre-based care and nannies. Informal childcare is care provided by neighbours, family members and friends. Mixed care refers to childcare packages that involve concurrent use of formal and informal care.

their childcare fees (Australian Government. Productivity Commission [P.C.] 2014, p. 383). As a result of these government subsidies the childcare costs of low income4 single parents in Australia are lower than those of single parents living in many other Anglophone countries (except for the UK). However, low income Australian single parents still experience substantially higher childcare costs than single parents living in Demark, Sweden or France (OECD 2014). Thus, while the average cost of childcare (after the subsidy) is substantially lower than in many other Anglophone countries, Australian parents seeking toddler and infant care still face high costs (Mahon et al. 2012).

In the last decade on-going criticism of the relatively high cost of childcare in Australia has been joined by new criticisms regarding its inflexibility and inadequacy for addressing the reality of Australia's 24/7 economy (Brennan and Adamson 2014; P.C. 2014). This criticism stems from the nature of Australia's childcare subsidy system. Most subsidized forms of childcare are relatively inflexible, whereas more flexible forms of childcare are difficult to access because places in these services are either tightly capped or not subsidized. Specifically, while parents who use registered childcare centres and registered family daycare can access government subsidies many of these services do not meet the needs of the large number of parents who work non-standard hours (Brennan and Adamson 2014; Craig and Powell 2011).The usual operating hours of childcare centres and family daycare is 7:30/8 am to 6 pm and families are generally required to book a regular day of care and not allowed to change these from week to week, though some family daycare educators do offer care outside of standard centre hours (7 am to 6 pm) and are more flexible about changing the days of care (Brennan and Adamson 2014).The only service specifically designed to meet the needs of parents who work non-standard hours is the Australian Government's *In Home Care* program which allows single and partnered parents, whose work shifts include hours during which no approved childcare service operates, to apply for access to subsidized in-home care (Australian Government: Department of Education 2012). However, not all eligible parents can access this program because the number of places is capped (Brennan and Adamson 2014). Within this social and policy environment informal carers, in the context of a mixed childcare package, play a critical role in helping single mothers to enter and sustain employment.

#### **The Study**

#### *Data Collection*

The findings presented in this chapter are based on interview data from a 3 year longitudinal qualitative study with 30 single mothers. At the time of the first interview all these mothers had at least one child aged less than 7 years, were living in Perth, Western Australia, and were in receipt of Parenting Payment, which is an

<sup>4</sup>Earning 50% of the average wage.

income support payment for single parents.5 Each mother participated in an annual interview lasting 50–90 min. Only one participant did not complete all three interviews. Half of the 30 single mothers in my study left income support during the 26 months of my research (a third due to re-partnering and two thirds due to being over the means test threshold as a result of their earnings from paid work). This is consistent with research that suggests that only 15% of single mothers in receipt of income support remain on payment continuously over a 4-year period (Barrett 2002; Gregory and Klug 2003).

Although each interview had a slightly different focus, all of the interviews explored the women's experiences of income support receipt, activation programs, participation in employment/study and use of childcare.6 The sampling strategy sought to select a group of women who captured the diversity amongst the population of single mothers with young children who are in receipt of income support. Potential participants were recruited using multiple strategies including attendance at a teenage mother's group, university press releases, an interview on local radio, the placement of posters in places commonly frequented by single parents (including shopping centres, medical clinics, indigenous women's health centres, childcare centres, and libraries), and finally placement of messages on single mother/parent e-lists and parenting web-boards.

#### *The Sample*

Interviewees' ages and number of children largely reflected the population of single mothers with young children. Interviewees were aged between 20 and 41 years, and had an average of 1.5 children. Most (two thirds) had only one child, while 13% had three children, and 20% had two. The sample was evenly split between those who had very young children (new-born to 3 years) and those who had an older youngest child (aged 4 to almost 7 years).

Mothers were similarly diverse in terms of their educational attainment, labour force status, and childcare packages. One third of interviewees had obtained a Bachelor degree (or higher) but slightly more (40%) had not completed high school.

<sup>5</sup>At the time of the first interview Parenting Payment (single) was available to all primary carers of a child aged less than 16 years who met strict income and assets tests. For full details of how the income and assets test of parents are taken into account when determining eligibility for Parenting Payment see Australian Government: Centrelink (2005). This booklet also outlines the associated allowances and payments these parents receive include concession cards, a pharmaceutical allowance, Family Tax benefit and a range of other small benefits.

<sup>6</sup> In terms of income support policies, the single mothers that are the subject of this article were required to seek 15 h of paid work per week once their youngest child turned 8 years old (Grahame and Marston 2012) However, if registered formal childcare was not available, or travel time was excessive, or if the parent would not benefit financially from the paid work they were not obliged to accept the employment offer.

Comparison with population figures reveals that single mothers who have a degree are over-represented in the study while single mothers who have completed high school but have no further education are under-represented (ABS 2008). The employment rate amongst the sample also appeared to be slightly higher than the broader population of single mothers with half of the interviewees reporting participation in some form of employment at the time of the first interview (ABS 2008). Given that eligibility for income support (Parenting Payment) is subject to income tests, it is not surprising that only two interviewees were employed full-time at the time of the first interview, while the remainder were working part-time on either a permanent or casual basis. In terms of childcare packages, at the time of the first interview interviewees commonly used mixed childcare packages (13 out of 30). Parental only care (n = 8), formal only (n = 6), and informal only (n = 3) were less common packages of care.

The analysis takes differences between interviewees' characteristics and the characteristics of the broader population by examining how a whole range of individual characteristics and institutional constraints work together in particular cases, rather than attempting to generalize in a straightforward way from the frequencies in the sample to the population.

#### *Data Analysis*

Interviews were fully transcribed before being analysed using Spencer and Ritchie's 'Framework' approach to inductive thematic analysis with NVivo 10. The Framework approach involves creating summaries of the individual *themes* for each *case* (interviewee). This involved: (1) immersion in the data through close reading of the interview transcripts; (2) developing the coding index through identifying the *themes* related to interviewees' experiences of employment that would form the Framework; (3) coding the data in NVivo 10 by systematically applying the coding index to the transcripts; (4) charting: creation of a 'Framework Matrix' in NVivo 10, with thematic headings and summaries of these themes for each case in order to "build up a picture of the data as a whole"(Ritchie and Spencer 2002, p. 317); and (5) mapping and interpretation: where the researcher "searches for patterns and connections" across the themes "and seeks explanations for these internally within the data" (Ritchie and Spencer 2002, p. 321).

Drawing on McKie et al.'s (2002, p. 910) concept of *caringscapes* I focused on themes related to the dynamics of experiences of care over time, and "the significance of the time and space-bound social relations inherent in the existence of locales". Furthermore, drawing on the lifecourse approach I focused on key transitions and turning points: (1) entering employment; (2) balancing work and care day to day; and (3) changing childcare needs across the lifecourse. Once it was clear that informal carers played a key role in helping single mothers to move into employment and sustain it over time, I then sought to identify individual case studies that illustrated "contrasting dynamics" (Shirani and Henwood 2011). This chapter

primarily focuses on specific cases that illustrate how these broader patterns of difference played out. That is, while the chapter makes reference to the patterns that were evidenced across the sample of 30 mothers; in the presentation of findings I primarily focus on specific cases so as to illuminate 'the complexities of "individual" families', and the "significance of the time and space-bound social relations inherent in the existence of locales"(McKie et al. 2002, p. 910).

#### **The Role of Informal Carers in Supporting Single Mothers' Transition from "Welfare to Work"**

Current childcare policy largely "does not incorporate an awareness of the demands of the everyday across the lifecourse" (McKie et al. 2002, p. 897). The result is that most single mothers who are employed struggle to make work and care align over the long term. Millar and Ridge's (2009) recent research on single mothers leaving income support highlights the important role that social relations play in enabling them to achieve employment sustainability. Specifically, their findings highlight the important role that social relations within (and outside) the family play in helping them maintain employment over time (Millar and Ridge 2009). This chapter elaborates on their findings by arguing that for women with young children a very specific form of social support, namely informal childcare from friends or relatives, in the context of a mixed childcare package, is crucial for enabling them to achieve strong employment trajectories.

My comparison of cases over time revealed that the relationship between the type of childcare package that a mother used and her employment trajectory emerged from the differential abilities of these packages to bring together, in space and time, the different locations within which childcare and paid work occurred (McKie et al. 2002). Interviewees who relied entirely on formal childcare (centre-based or family daycare), or entirely on informal childcare, experienced greater struggles with moving beyond low part-time hours, or transitioning into higher quality employment, compared to interviewees with similar characteristics who used mixed childcare packages. In the following it is argued that many of the difficulties experienced by those using only formal care emerged from difficulties with finding care that adapted to children's and parent's changing needs, and which fitted with their employment circumstances. Single mothers who used mixed childcare packages (formal childcare and informal care concurrently) experienced much stronger employment trajectories as a result of (1) a greater ability to obtain suitable care as soon as they received an offer of employment or started a new program of study; (2) a greater capacity to balance work and care on an on-going basis; and (3) an ability to fall back on informal care when their children's needs or their employment arrangements changed, thereby rendering formal arrangements no longer suitable.

#### *Entering Employment or Education, and the 'Helping Hand' of Informal Care*

A single mother's entry into employment is contingent upon finding childcare that matches her work hours. Typically she must commence work soon after an employment offer is made, but locating a formal childcare placement for the required hours/ days commonly takes weeks or months. The presence of kin (or friends) who are willing to provide childcare until she can find a suitable formal childcare place may thus be critical to her successful transition into paid work. A typical example of this kind of experience was that of Sandra who worked as a part-time lecturer on short, fixed-term contracts. Sandra finished her PhD shortly before the birth of her first child and when her daughter was 4 months old she was offered a position as a parttime lecturer but was unable to find a place at local daycare centres. Sandra explained:

I mean in the first semester I worked I couldn't get [Rose] into a daycare so she had to go to family and it was only that second semester that they [the childcare centre] gave me a half a day a week; that is all they could give, and I just had to keep taking her in for half a day so that I could get the days I needed.

*[ ]* = *added text*

By the time of the second interview, Sandra had increased her work hours and was working full-time across three separate part-time appointments; and as I discuss below, her mixed childcare package also assisted her with the challenge of balancing work and care on an on-going basis.

Given the reality of long waiting lists for most childcare centres, single mothers may not be able to accept an offer of employment if they do not have kin who can provide short term care until they locate a place in formal childcare. Linda's experience was an illustrative example of this problem. At the time of the first interview, Linda had very recently separated from her husband, and she had primary custody of their three children (aged 4–13 years). She had comparatively low labour market capital because she had engaged in little paid work since the birth of her first child 13 years prior. Her ability to access informal care was highly limited due to her social isolation. Her separation with her ex-husband was acrimonious with the result that no regular care arrangements could be negotiated with him over the 3 years of my study. Furthermore, all her immediate family lived in another state and while Linda had acquaintances through her children's activities, she had few close friends. Despite these multiple barriers to paid work, by the time of the second interview she had obtained three regular casual shifts as a waitress. All the shifts involved working 1 or 2 hours outside school hours, and during these periods she relied on her oldest child to care for the two younger siblings. Linda was keen to obtain employment with better pay and conditions, and in the period between the first and second interview, she applied for and was offered a full-time secretarial position with better pay. Unable to find before and after school care at short notice, Linda again relied on the oldest child to care for his siblings. However, this time he was responsible for helping get them ready for school and heating up dinner five times a week. During her 1 week trial of her new job Linda decided that (in her own words) "it just wasn't workable." She quit and returned to her low paid waitressing position. We may view this as being a "critical moment" (Giddens 1991, pp. 112–3; Holland and Thomson 2009; Thomson et al. 2002) for Linda where things came together in a way that she had to make a decision that would significantly shape her employment trajectory over the medium term. In this case Linda's inability to find suitable childcare at short notice amplified the stress associated with returning to full-time work while continuing to caring for three children. Although the lack of informal support with childcare was not the single factor that determined her decision, it is clear this absence reduced her ability to take up this new position.

#### *Balancing Work/Study and Care*

Single mothers' abilities to balance work and care, or formal study and care, on an on-going basis was shaped by a number of factors; including their personal feelings about the appropriateness of using non-parental care for young children, their work/ study schedules, and their childcare package. Sandra, whose story of entering employment following her daughter Rose's birth was outlined earlier, continued to use mixed care in the second year of my study. She explained Rose's childcare schedule as follows:

On Monday she stays with her dad's mum. Tuesdays she's at daycare all day. Wednesday morning she's at daycare, then she's with her dad in the afternoon. And Thursday all day she's been at daycare just because [this year] my parents are doing the grey-haired nomad thing, and travelling around Australia in a caravan. And on Fridays she's been at her dad's mother's.

*[ ]* = *added text*

When I asked why she used this care option she explained, "Well, daycare is expensive" but also she explained that she was unwilling to place Rose in formal childcare for more than 2 days a week. Thus, when I asked Sandra, "What would you do in terms of work, paid work, if you didn't have the help from your family?" She replied:

I don't know, you know. I think I'd be quite stuck…sometimes I wish I wasn't a solo parent. But, other times I think I've got it a lot luckier than other women that I know, who don't have that support network. And, don't have a father who's interested in fathering the child, you know. Like I really feel like that [not having support from family] would be really hard. *[ ]* = *added text*

One of the common cause of clashes between paid work and care obligations is unpredictable work-schedules, and working non-standard hours. Consistent with this, interviewees with these work schedules experienced more challenges in balancing work/study and care on an on-going basis. Consistent with existing research (Williams et al. 2013), highly educated interviewees in professional employment usually had predictable work schedules, worked within standard business hours, and when they did work non-standard hours they had some flexibility in their schedule. As Nina, a technologist at a major hospital, explained, "I've had to…change my shift at [hospital name] from an evening shift to a day shift …I just swap …there is a few people." Interviewees with low education in non-professional employment in contrast experienced what I call *erratic rigidity*. Not only did they commonly have to work outside standard business hours, and have schedules that changed from week to week, they also usually had very little ability to influence the schedules they were assigned (Williams et al. 2013). For interviewees who used mixed care, the challenges associated with these kinds of schedules were significantly reduced. An illustrative example is the experience of Julie, who commenced full-time employment as an events manager between interviews one and two, and who had three children, including a daughter who was only 6 months old at the time of the first interview. Julie explained that she had to work until an event finished which meant that:

at least once a week [I work late at night], sometimes twice, not usually until two, but usually twice a week until 10.30 till 11… So I work nights, I work weekends.

Such experiences are common among Australian single mothers with almost half (45%) reporting they have no say in their start and finish times (ABS 2009). Employed single mothers' concentration in non-professional employment in part reflects the relatively low average level of education amongst this population (ABS 2008).7 In Julie's case her use of a mixed childcare package helped her to balance work and care. Her package included centre-based care for her infant daughter, and after-school care for her sons during the week, while most of the weekend they stayed with their father (Thursday, Friday and Saturday night). Julie explained at interview two that she had relied a lot on her family to balance work and care after she had taken up her position as an events manager, because she had not realised "how much night work and weekend it [her job] was going to take". For a period of a few months she relied on her mother to care for her infant daughter, and also on her wide network of close friends. In the longer term this heavy, on-going reliance on family and friends was not sustainable and Julie eventually obtained in-home care. However, in the medium term, her family and friends provided essential assistance with balancing work and care.

#### *Managing Changing Childcare Needs Across the Lifecourse*

Children's childcare needs change rapidly as they grow "and, as careers/jobs change, so do the needs of parents" (p. 910). However, as McKie et al. (2002) have argued, this obvious point is largely ignored by national childcare strategies and government funded programs and services which overwhelmingly "concentrate on younger children and ignore lifecourse changes" (p. 910). The highly dynamic nature of children's and single parents' childcare needs was evident across the 3 years of my study.

<sup>7</sup>Two in five Australian single mothers have not completed high school, and only 15% have a degree or higher.

Children grew, their needs changed, and thus they needed new childcare arrangements. As one mother put it when I asked if her daughter was still attending family daycare now that she had started school, "My daughter out-grew family day-care." These kinds of changes were very common and almost all the mothers who used formal childcare reported at least one instance of their child 'out-growing' or suddenly forming a dislike for their formal childcare. In some instances, this apathy arose when the child was moved to a new childcare centre or when they were moved up to a different room/age group within the same centre; whilst in other cases these problems appeared to be simply a consequence of the child growing older and developing new preferences. An illustrative case of the former problem was Sophie's experience of her toddler son [Caylen] refusing to attend childcare following his move to a new centre as a result of her moving to a new college. As Sophie explained:

When I was going to [first college] they had an adjoining daycare centre. He loved it at first, he had a friend that he just loved and the care was better. But then I moved to a different college and he had to move to a different daycare he started having troubles…I got worried because he was objecting so strongly that there was a situation or something that was happening. I got worried about that. Maybe it was too strict, maybe they spoke to him in a way that made him feel bad…I really struggled with it, if I should keep doing that. It was really breaking my heart to leave him like that. It was awful.

Concerned that forcing Caylen to continue to attend would have a detrimental effect on him, Sophie reluctantly make the decision to drop out of college and to defer further study until Caylen was a bit older. Three years later Sophie decided to commence part-time studies. As the weekly class was scheduled in the evening (6:30 to 9:30 pm) she relied on her parents (primarily her mother) to provide care for Caylen. When I asked why she had not relied on her parents for care earlier, she reported that she was unwilling to ask them to provide that much care. The following year, when Caylen began pre-school, Sophie commenced part-time work (2–3 days a week) at a bookstore during school hours in addition to her part-time studies.

Rebecca's experience, in contrast, illuminates how access to informal carers in the context of a mixed childcare package helps support a stronger employment trajectory. When I first interviewed Rebecca, she was completing a Bachelor of Social Work, and had one daughter (Violet) who was 2 years of age. Rebecca frequently relied on her parents to care for Violet and also supplemented this with 2 days of care at a childcare centre. Thus when her daughter Violet starting objecting to attending childcare she relied on her parents until she could make alternative arrangements. As Rebecca explained at the second interview:

she [Violet] loved when she was in the baby room but when she was two and a half they bumped her up to the big kid's room and she didn't like it, she didn't get the same attachment to the teachers or the other kids. …..At the [centre-based] daycare Violet was crying not to go, not to be left here, and she wanted to go back to the baby room, and if one of those teachers came through the room during the day she would get upset.

Rebecca persevered with keeping Violet at the centre "for a couple of months because it was just a couple of days a week." However, when she "didn't change" then Rebecca's "mother looked after her [Violet] until [she] was able to find a home

[family] day care." During this period Rebecca continued with her studies, and by the third interview she was working in a position related to her degree.

#### **Discussion and Conclusion**

Existing studies demonstrate the important role that access to high quality, affordable childcare (Scott et al. 2005), and social support (Cook 2012; Millar and Ridge 2013; Ridge and Millar 2011), plays in supporting this transition. However, many dimensions of the relationship between childcare and single mothers' engagement in paid work have been overlooked. Within this chapter I have drawn on the concepts of the lifecourse and caringscapes to analyse data from a QL study that sought to understand the work and care experiences of single mothers with young children who were in receipt of income support. This approach has allowed me to make some important contributions to the current literature on single mothers.

Firstly, by taking a lifecourse perspective and drawing on a longitudinal study, I have revealed the degree to which the childcare needs of single mothers and their children are dynamic. This has been largely ignored in the literature which tends to emphasize the need for single mothers to obtain high quality, affordable care at a point in time. The implicit assumption underpinning this approach is that obtaining childcare is a one-off event. However, within this chapter I have illustrated that obtaining childcare involves single mothers engaging in on-going work to find childcare in a context where children's needs change and so do parents', and formal childcare is quite inflexible.

Secondly, this chapter makes a significant contribution to the literature on single mothers' childcare needs. To date the literature has emphasised these mothers' needs for affordable and high-quality care (Brady 2010; Cabrera et al. 2013; Scott et al. 2005). The findings presented here suggest that we also need to consider the issue of childcare inflexibility. While the problem of inflexibility has been briefly mentioned in the literature on mother's use of informal carers (Backett-Milburn et al. 2008; Millar and Ridge 2009), the literature has not sought to systematically identify the particular areas in which current childcare systems are inflexible.

This chapter suggests that there are four key ways that the current formal childcare system in Australia (and likely elsewhere) inhibits single mothers seeking to complete education or move into paid work, and to maintain a strong employment trajectory over time. Firstly, long waiting lists for formal care, particularly for very young children, frequently prevent single mothers from locating suitable childcare when they receive an employment offer or commence a new semester of study. The result is that while those who have family or friends available and willing to provide care may turn to them to provide "stop gap" care until a formal care place can be located, mothers who do not have this assistance will most likely have to decline the employment offer or withdraw from their study.

Secondly, the sharp age segmentation within current childcare services forces families to find different carers for children of different ages, and to change carers every few years as their children grow. Single mothers who had multiple children and access to informal carers commonly obtained their assistance with picking up (or dropping off) children at multiple locations. Mothers with multiple children who lacked access to informal care commonly struggled.

Thirdly, there are few formal childcare options available when single parents require emergency care because their usual arrangements have fallen apart. Informal carers may play a critical role in helping single mothers to adjust to these changing circumstances. A fourth and final dimension of inflexibility is the significant shortage of formal childcare options for single parents working non-standard hours. Currently there are very few twenty four hour childcare centres, and the *In Home Care* program is poorly advertised and also has limited places (Brennan and Adamson 2014).

What, then, are the practical implications of these findings? The first is the need for much more sustained attention to the problem of childcare flexibility. Existing statistical studies (Baxter et al. 2007; Brady and Perales 2016), and the analysis in this chapter, find that single parents who use mixed childcare have greater labour market success; because a mixed childcare package is more flexible than formalonly or informal-only childcare. In other words, the current childcare system makes it hard for single mothers who cannot access mixed childcare to succeed in the labour market. There is a need to develop a more flexible system that enables parents who have a personal preference for mixed care to continue to use this childcare package, but which does not disadvantage single mothers who cannot access a mixed childcare package. In Australia the issue of childcare flexibility is just starting to gain more attention from policy makers. In 2013 the Labor Government launched the 'Childcare Flexibility Trials' to trial more flexible care options, but the evaluation found that take-up of the trialled options was very low (Baxter and Hand 2016). A year later (2014) the new Coalition Government asked the Australian Productivity Commission to conduct an inquiry into childcare, and one of the six terms of reference specifically focused on options to improve flexibility (P.C. 2014). However, the pilot of subsidized nannies that was recommended by the commission has also struggled with poor take up (NewsABC 2016). This chapter suggests that further research is needed to more fully understand how the formal childcare system can be made more flexible, so as to address single mothers' needs when combining caring and paid work across the lifecourse. Such reforms are necessary to move beyond viewing the issue of work-life balance in terms of individual solutions (Brady 2008).

**Acknowledgements** I would like to thank participants at the Lone Parenthood in a Life Course Perspective workshop, 6th–7th June 2014 – University of Lausanne, and the two anonymous referees, for constructive feedback on earlier versions of this chapter. The field research upon which this chapter is based was supported by funding from a Provost Doctoral Entrance Award and a Department of Sociology Doctoral Research Grant, both from the University of Alberta.

This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **References**


**Michelle Brady** is a Senior Research Fellow in Sociology in the School of Social Science, University of Queensland. Prior to joining UQ in 2012 she worked as an Assistant Professor at the University of Victoria, Canada. Michelle received her PhD in Sociology in 2011 from the University of Alberta. Currently she is the lead Chief Investigator on an Australia Research Council Discovery project on childcare flexibility. She is also lead author on offical evaluations of Australia's first publically funded paternity leave scheme, and in-home childcare scheme. Her work has been published by the Journal of Social Policy, Work, Employment and Society, University of Toronto Press and Critical Policy Studies.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Chapter 12 Lone Mothers in Belgium: Labor Force Attachment and Risk Factors**

**Emanuela Struffolino and Dimitri Mortelmans**

#### **Introduction**

The risk of welfare dependency for lone mothers is one of the major concerns in the socio-political debate connected to the increasing number of lone parent families in most of the European countries. In fact, lone mothers are more likely than mothers in couples to be poor, unemployed, and welfare recipients. Even though paid work is considered one of the preferential ways to avoid poverty and welfare dependency, individual and family characteristics define a variety of labour supply strategies. It follows that the typical work-first and welfare-to-work measures aiming to support employment for lone mothers that have been implemented in many countries might actually target a group that is highly internally differentiated, thus resulting in extremely mixed outcomes (Haux 2013; Lewis 2009).

Lone parents are mainly women, and as such might be confronted – when the transition to lone parenthood occurs – with additional burdens due, for example, to difficulties in re-entering the labour market after inactivity or in existing social assistance because of the increase in care responsibilities driven by being primary carer and breadwinner. It follows that experiencing lone motherhood might result in strategies for balancing work and life by mixing welfare and labour supply that differ according to previous and current resources as well as household characteristics. Furthermore, changes in the demographic characteristics of lone parents (Berrington 2014; Kiernan 2004) are likely to interact with traditional factors of inequality once one looks at specific life domains, such as integration within the national/regional labour market or through the welfare system itself.

E. Struffolino (\*)

D. Mortelmans CLLS, Faculty of Social Sciences, University of Antwerp, Antwerp, Belgium

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_12

WZB Berlin Social Science Center and Humboldt University, Berlin, Germany e-mail: emanuela.struffolino@wzb.eu

The relationship between lone parenthood and labour market participation has been studied in terms of (i) the differential probability of being employed, comparing lone mothers and mothers in couples (Western et al. 2008); (ii) single transitions (and their timing) in and out of the labour market after becoming a lone mother (Stewart 2009); and (iii) the responsiveness to incentives (if any) for lone mothers to move off welfare and into work (Athreya et al. 2010; Dowler 1998; D'Souza et al. 2008). However, the strategies adopted to balance family and work domains after the transition to lone parenthood cannot be captured by means of cross-sectional analysis looking at behaviours at a single point in time over the life course.

Women who experience lone motherhood for a certain period of time over their life course have to face the resulting increase in economic and care needs in the household. Despite the tremendous changes that have occurred in terms of women's employment opportunities and configuration of gender roles within the family, women are more exposed than men to poverty risk due to the decrease in household income after separation or divorce (Andreß et al. 2006; De Regt et al. 2013; Jarvis and Jenkins 1999; Poortman 2000; Smock 1994; Smock et al. 1999). The strategies individuals can adopt to respond to the consequences of changes in the family life domain, such as increased economic and care burdens, depend on the characteristics of both the welfare system and the labour market structure.

By embracing a life course perspective and thus considering the dynamics of family trajectories, we stress the importance of looking at differentiation within the group of women who experience at least one episode of lone motherhood right after their break-up. In fact, it has been argued that family disruption is one of those new social risks that might shape differently the exposure to vulnerability of different groups when interacting with more traditional social risks, such as a weak labour market attachment or welfare dependency (Bonoli 2005; Vandecasteele 2010). For this reason, looking directly at the variety of experiences of this specific group of women who have to deal with work-family reconciliation to a much greater extent than mothers living in couples might shed light on the effect of household characteristics and the timing of the transition to lone motherhood on labour market participation and/or welfare dependency (Drobnič 2000).

Building on the few previous studies on the variability among individuals' employment trajectories following the transition to lone parenthood (Stewart 2009; Zagel 2013), the aim of the present chapter is to describe the nature of such variability in the Flemish case. In Belgium, lone parent families represent almost one fourth of households with children (Wagener et al. 2014). Regional differences exist, but the proportion of lone parent families exceeds at least 20% everywhere. In Flanders in particular, the poverty risk for this type of family is significantly higher compared to couple-with-children households and to the general population itself, even though welfare provisions are available to a relatively great extent for different types of recipients. Furthermore, on the one hand, the Flemish region accounts for the majority of women who exploit the career break system meant to encourage labour force attachment and return after parental leave, but on the other hand, it scores worse than the other Belgian regions where female labour market participation is concerned.

In this context, the high incentives to take employment and the low impact of childcare costs on income gain are also combined with relatively generous unemployment benefits for the long-term unemployed (Immervol and Scarpetta 2012). These factors might lead to opposite outcomes when the labour supply strategies of lone mothers are considered. In fact, two opposite mechanisms can affect labour market participation: on the one hand, favourable childcare arrangement and further incentives for working can facilitate it, and on the other hand, welfare generosity might be discouraging. Nevertheless, from a medium and long-term perspective, a weak labour market attachment can undermine future opportunities due to skills depreciation processes. Since the result of this trade-off between the needs for care and for income within the household that emerges during lone parenthood is shaped by both individual and household characteristics, inequalities within the population of lone mothers might reinforce existing disadvantages in terms of differential in the risk of social exclusion driven by family structure and labour market arrangement. Thus, we firstly consider whether the age at the transition to lone motherhood acts as a stratification variable when looking at labour market opportunities, which are frequently path and seniority dependent. Furthermore, we explore the relationship between employment trajectories and partnership status before lone parenthood, as well as the number and age of children in the household as a proxy for the potential effect of care responsibilities on labour supply. Register data from the Crossroads bank of Social Security (CBSS Datawarehouse) allow us to track labour market participation pathways of Flemish women who experienced the transition to lone parenthood after divorcing in 2004.

#### **Lone Parenthood Between Welfare and Labour Market Participation**

Partnership dissolution can be seen as a risk not contemplated as part of standard life-course expectations as conceived in the industrial society (Vandecasteele 2011). The increase in the share of individuals who experience lone parenthood is thus a by-product of the second demographic transition, and increasing exposure to the economic risks of this type of family is a consequence of the transition to a postindustrial society, which is associated with changes in women's labour market participation, growing unemployment and in-work poverty. Against this scenario, the role of coupledom in providing life-long economic security is revealing its inadequacy. The transition to lone parenthood is accompanied by changing needs in terms of care and income that can trigger employment choices strongly shaped by resources made available by the welfare state. It follows that life chances and exposure to what are considered new social risks connected to life-course events in the family domain are strongly shaped by welfare state provisions and labour market structure (Bonoli 2005; DiPrete and McManus 2000; Taylor-Gooby 2004).

When considering the relationship between family structure and employment choices, one has to acknowledge that different institutional settings provide lone mothers with different scenarios they might access, since welfare state regimes can indeed either promote a traditional male breadwinner model of work and care or rather encourage dual-earner families. In both cases, however, if welfare support is scarce, the risk of poverty increases as a consequence of the sudden lack of money from the father and necessitates a re-arrangement of labour supply for the mother, who usually holds custody of the children and thus the increased need for care within the household (Christopher et al. 2000; Esping-Andersen 1999; Garfinkel et al. 1994).

Cross-national studies show that lone mothers work less than mothers in couples, and that lone mothers' full-time engagement in the labor market is higher in countries in which work policies are more flexible (Plantenga et al. 2010; Ruggeri and Bird 2014). In other words, even if in many countries the activity rate of lone mothers has been increasing during the last decades (Cohen 2002), not having a partner exposes women to higher unemployment risk and persistent poverty (Eamon and Wu 2011). The process of feminization of poverty has been acknowledged since the 1990s for the United States (Daly 2000), and this is likely to shape the new borders of poverty risk in many European countries in correspondence to the increasing share of women who experience lone parenthood. In fact, women suffer more than men from the decrease in household income after divorce (De Regt et al. 2013; Jenkins 2009; Rake 2000), with recovery from this drop in economic resources lasting between 4 and 6 years depending on the context (De Vaus et al. 2014; Mckay 2002; Mortelmans 2017). Furthermore, experiencing exposure to poverty risk, as well as to stress factors caused by the trade-off between care and employment during periods spent as lone parents, has been found to become chronic even after repartnering (Friedland and Price 2003).

When looking at the interaction between welfare, employment, and lone parenthood in contexts in which self-sufficiency and economic independence are rather strongly promoted by activation policies for benefit recipients (Daly 2011), the main issue lone mothers have to face is how to combine work with parenting (Haux 2013). In the literature, the concept of "barriers to employment" formalizes the pool of factors interacting in such reconciliation process (Moffitt et al. 2002). Both individual and structural characteristics might represent barriers: in fact, personal physical/psychological health, individual skills, and work experience can act both independently from and in interaction with additional obstacles such as family size, availability of affordable childcare (Daycare Trust 2011; Millar and Ridge 2008), employer prejudices and, in general, labor market opportunities (Gingerbread 2012) 1 in threatening the possibility of increasing one's labour supply.

<sup>1</sup>Several analyses cannot account for individual preferences to stay at home to care for children vs. working (Tomaszewski et al. 2010; Duncan and Edwards 1997): not controlling for this unobserved heterogeneity for the selection into unemployment might lead to misinterpret the (spurious) relationship between actual employment status and lone parenthood. Nonetheless, it is worth firstly noticing that evidence on impact of individual attitudes and institutions is mixed: on the one

Consistently with the aim of this chapter, we focus here only on some of these factors. The first structural characteristic is the welfare system. On the one hand, policy measures for supporting lone mothers can sustain employment, such as in the cases of Germany and the US, where public childcare is made available for lone mothers, who work at higher rates than mothers in couples (Zagel 2013). On the other hand, policies can promote a social assistance payment system, as in the United Kingdom where in general lone mothers work less than mothers living in couples (Millar and Rowlingson 2001). Income-related arrangements (such as welfare subsidies for poor families, which are usually lone parent families) are found to be more effective in alleviating the economic drawbacks of divorce than employmentrelated arrangements (Uunk 2004) and to promote employment especially among low-income lone mothers (Ahn 2012). The availability of reconciliation policies itself might not be sufficient to eliminate the barriers to work. For example, long childcare leaves without benefits are suitable for parents in a family with one breadwinner, while widespread public or subsidized childcare support reconciliation for more family types and women's labour market participation in general (Boeckmann et al. 2014; Korpi 2000; Korpi et al. 2013; Van der Lippe and Van Dijk 2002). In fact, high costs for childcare are discouraging, especially for low-income families and lone parents in particular (Connelly and Kimmel 2003). When a family cannot rely on informal help from relatives, then the market supply of working mothers might decrease to a great extent (Herbst and Tekin 2011).

However, assuming that lone mothers react to the welfare state configuration by following a rational economic approach, as is more or less implicitly expected by many welfare-to-work measures, would be misleading. In fact, not just because plenty of opportunities for working full time are available, lone mothers might decide to fully delegate childcare (Duncan and Edwards 1997). In other words, gender pressure might act independently from economic pressure. Furthermore, the care-employment trade-off with which women have to cope is strictly connected to both the different social expectations towards motherhood and the type of labour demand lone mothers can actually access. In fact, in several countries in which the male breadwinner model prevails, mothers in couples are not expected to be strongly engaged in the labour market due to their care responsibilities within the household. In those same national contexts, lone mothers are instead asked to increase their labour supply to contrast welfare dependency through work. Evidence shows that lone parent families rely more and for longer periods on welfare support than other types of household do (Voges et al. 1996). However, as the increase in in-work poverty shows clearly (Andreß and Lohmann 2008; Cooke and Lawton 2008), working itself does not protect against poverty, being a necessary but not sufficient condition for self-sufficiency. In the presence of low wages and unfavourable working conditions (temporary contracts, shift works, no health insurance, no vacation benefits, etc.),

hand, structural barriers rather than preferences are found to be crucial, and especially those connected to childcare and labor market opportunities (Whitworth 2013); on the other hand, there is some evidence that shows that those who experience lone parenthood at certain point over the life course had already low rates of employment (Mckay 2002).

lone mothers might not be able to move off welfare through work, and thus they have to combine welfare and work (Cancian and Meyer 2000; Eamon and Wu 2011; Millar and Ridge 2008; Wu et al. 2008).

Structural barriers in particular interact with individual and household characteristics. Where the first are concerned, previous research shows that when motherhood occurs early in the life course, individuals who did not complete education could be funnelled in a spiral of cumulative disadvantages in the labour market and wage-wise as well as across life domains (Chevalier and Viitanen 2003; Looze 2014). Furthermore, it can be assumed that when lone parenthood has occurred at a young age, mothers might not have enough qualifications and labour-market experience to endorse and thus have less bargaining power and are more exposed to underemployment. In contrast, those who become lone mothers at an older age have cumulated experience, and this seniority benefit (and related benefits) might increase the opportunity to obtain better compromises to reconciliate paid work and family care (Ondrich et al. 2003). Similarly, older lone mothers are found to be less likely to suffer from economic hardship (Bauman 2002)2 . Results from cross-sectional data of European countries as a whole show that young lone mothers work less than young mothers in couples, and that this gap is smaller across groups of older women (Ruggeri and Bird 2014).

The relationship status before lone parenthood – namely whether the couple was married or cohabited – cannot be strictly conceived as a "barrier" to employment, but there is evidence that some kind of selection results in cohabiters showing a more equal division of paid work compared to married couples, and married women being on average more economically dependent on their partners (Snoeckx et al. 2008). Thus, divorced women suffer from a greater disadvantage for being divorced than cohabiting women who were formerly more independent, thanks to a stronger presence in the labour market. It follows that heterogeneity exists in the economic effects of union dissolution according to previous relationship status, but it is mainly driven by previous employment status. In this understanding, having been married or having cohabited before the transition to lone parenthood might represent an additional disadvantage connected – again – to experience in the labour market and to age at lone parenthood.

When considering household characteristics, the number and age of children have been found to be crucial in allowing lone parents to give up employment and care for their children as their main occupation (Kasparova et al. 2003; Millar and Ridge 2008) . For example, in Germany, mothers in couples are less likely to work

<sup>2</sup>Unfortunately, as discussed below, here we are using register data and information on education is not available. However, we know from the human capital theory (Becker 1975; Mincer 1974) that different investment on education defines different opportunities in the labor market and in wage growth pace. In particular, leaving labor market for some years is found to matter more for highly educated than for low-skilled women (Stewart 2009). It follows that since highly skilled women invested more in education and they expect higher returns, they will be less likely to leave the labor forces (and relying completely on welfare) and more likely to have external resources to mobilize for re-conciliating work and care independently from formal childcare availability and its cost.

than lone mothers, but when they have a larger family they re-enter the workforce (and often full time) to the same extent as lone mothers do. In the US, heading a large family interferes with work off welfare (Harris 1993), but the age of the youngest child in the household is even more important in defining the labour market supply of lone mothers (Drobnič 2000). Similarly to the US, evidence from the UK and Australia shows that welfare measures promoting employment of lone mothers are more effective as the children grow up, and lone mothers with children older than pre-school age have much higher employment rates (Doiron 2004; Ermisch and Wright 1995).

This brief review aimed to account for the available evidence on the association between lone motherhood and labour market participation. However, most of the research either relies on cross-sectional data (which are not suitable for looking at evolution of processes over time) or uses longitudinal data for estimating the timing of occurrence of certain events, such as re-entering the labour market or getting a certain type of job. Even though it can be argued that selection into lone parenthood exists, an increasing heterogeneity is found when observing different cohorts of births and the distribution of lone parents over variables such as education and age (OECD 2014). It follows that differences in the mechanisms behind labour market participation strategies could not emerge when looking at cross-sectional data, while what really differs is the trajectory/pattern according to different stratification variables. Most of the research reviewed above relies on cross-sectional data. However, when we consider lone parenthood, the time from the transition to the new family arrangement is crucial to understand how strategies to combine care and work evolve: cross-sectional analyses are blind to such changes because they cannot follow individuals over time. The little existing evidence from longitudinal data shows that activity rates among mothers (including welfare recipients) are higher than the results of cross-sectional analyses (Harris 1993), and studies that exploit longitudinal data to look directly at trajectories of lone mothers show great variability according to the duration of lone parenthood (Zagel 2013; Stewart 2009). These results support the idea that bringing about within-group comparisons is useful to detect grey areas where old social risks and new ones connect to household structure and where the labour market and welfare system intersect.

#### **Lone Parenthood in the Flanders**

Belgium's history is characterized by remarkable regional developments, as it consists of a Catholic northern part (Flanders) and an atheist southern part (Wallonia and Brussels) with different spheres of influence. While Flanders belonged to the German part of Europe, Wallonia resided in the Latin part. After the independence of the country, the linguistic borders always remained a cultural demarcation line in unitary Belgium, resulting in important disparities with regard to secularization and demographic behaviour. The enlightened ideas stemming from the French Revolution were more widely spread in the southern part of the country than in

Roman Catholic Flanders. After World War II, the Belgian state structure evolved from a unitary state to a federal state, with central federal authority weakened and oversight of employment, education, and welfare transferred to two different types of regional authorities: communities and regions. Three communities – Flemish-, French-, and German-speaking – have jurisdiction over cultural matters, education, the use of languages, and matters relating to the individual, such as health policy (curative and preventive medicine) and social assistance (protection of youth, social welfare, aid to families, immigrant assistance services, and so forth). There are also three regions, which are geographically different from the communities: the Flemish region, the Brussels Capital region, and the Walloon region. These execute responsibilities with territorial impact, and their power extends to economic matters, employment, agriculture, water policy, housing, and public works.

With respect to Esping-Andersen's welfare state typology, Belgium is usually assigned to the group of Continental European welfare regimes (Esping-Andersen 1990). It has a low score in the cluster of liberal regimes and an average score in the cluster of social democratic regimes. In addition to having a high degree of decommodification, Belgium scores considerably high on "corporatism" and "etatism" indices. The system makes a strict division between regulations for employees, civil servants, and self-employed persons. A second trait is the degree to which health and unemployment benefits and (partially) pensions are related to earnings. This is a carry-over from the classic male breadwinner model, which is deeply rooted in the system. This principle is also represented in generous family benefits and social transfers, which support a traditional division of labour. Nevertheless, an elaborate childcare infrastructure, which enables women to combine work and family, provides evidence that Belgium also has characteristics of the dual-earner model.

Even though this chapter focuses only on the labour market transitions of Flemish lone mothers, we need to take into consideration the dual nature of the Belgian state structure. Divorce legislation as well as a large part of the labour market policy are centrally regulated, while economic prosperity and labour market participation clearly differ by region. The Flemish case is interesting because the relatively low female labour market participation is combined with a widespread use of "time leave" schemes. In Belgium, employees are entitled to interrupt their careers temporarily by means of the Career Break Scheme. During this period, employees are given a limited replacement income. The Flemish government encourages the use of the system by giving additional financial incentives. The majority (65%) of people in the career break system in Belgium are situated in Flanders. Several options are available to take a career break: full-time breaks, part-time breaks, and specific thematic leave. This schedule was meant to encourage longer stays on the labour market and would encourage returns to work after maternity leave. The system does not compensate the whole income from labour. As a consequence, lone parents are scarcely using the system to enhance their work-life balance.

Next to the lower participation in leave schemes, we also find considerable lower labour participation levels among lone parents. According to the Labour Force Survey, approximately 68% of all lone parents are working, although not full time. One third of lone parents (35%) work part time. We also observe higher percentages of lone parents in unemployment. Estimations of the Crossroads Bank of Social Security (CBSS Datawarehouse) show that the level of unemployment is five times higher among lone parents (32.2%) than among dual earner families with children (6.5%). Compared to the study of Geurts (2006), this is an increase of 14%. Again, it comes as little surprise that unemployment among lone mothers is much higher than among lone fathers. The demographic evolution of Flemish lone parents and its relationship with poverty risk and changes in income is extensively presented in Chap. 10 in this Volume. However, it is worth highlighting here that in Flanders, lone parent families show a higher risk of poverty compared to any other family type and to the general population (Defever et al. 2013).

Given the characteristics of the welfare system and the schemes connected to labour market participation, the Flemish case represents an interesting field in which to observe the evolution of employment trajectories after lone parenthood. In fact, typical patterns of labour/welfare supply might appear differently within the group of lone mothers and therefore strengthen old inequalities or rather generate new disadvantages.

#### **Data and Methods**

Register data represent an invaluable source for studying social processes because they are not affected by memory bias. The Crossroads Bank of Social Security Datawarehouse (CBSS Datawarehouse, released 2009) combines demographic data from the National Register with labor market and income data from several Social Security offices in Belgium.

We selected a subsample of women between 25 and 55 years old who separated or got divorced in 2004 after cohabiting or being married. The age of 25 was chosen for reasons of data availability. The number of broken marriages and cohabitations (with children) for ex-partners below age 25 was too limited to be included in the analysis. The age of 55 concerns labour market attachment, which weakens rapidly after 55, showing a different sociological reality that is not the focus of this chapter. We additionally restrict our sample to those who do not live together with a partner and who live with one or more (own or step) children younger than 18 or younger than 25 if they do not have a labour market income (Letablier and Wall 2017). We ended up with a sample of 2602 lone mothers3 , representing 34% of the original CBSS Datawarehouse sample of women aged 25– 55 in 2004.

For the first analytical step, we refer to the theoretical and methodological framework of sequence analysis, which considers individual processes unfolding over time – understood as a whole – as the primary unit of analysis. This approach to time-related phenomena is not grounded on any assumption about the generative process underlying the regularities in the sequence. In other words, causal modelling

<sup>3</sup>For a more detailed description of the sample identification procedure, please refer to Chap. 10 in this Volume.



only intervenes in a second step of the analysis because the first aim is to describe sequence structures and the relationships among them to try to discover typologies. One of the main advantages of considering sequences as a unit of analysis is that it fosters the description of social processes that empirically result in patterns of events (Abbott 2001, 2004). Only after the identification of the latter is it possible to consider the determinants of such patterns and their configuration, as well as whether and how they can shape other processes.

Thus we built individual sequences starting in 20034 (i.e. one year before the transition to lone motherhood occurred) and ending 5 years after the lone parenthood occurred. The original data were quarterly, so we ended up with a sequence of 24 quarters. Each time point was coded according to the position occupied by the individuals in or out of the labour market and the welfare system. In fact, in case of inactivity/unemployment, we define the individual position according to the welfare provision received. Table 12.1 displays the states used for coding the individual sequences.

Unemployment/inactivity benefits are available for those who are currently out of work, available to participate in the labour market, and living in Belgium. The minimum wage is a welfare provision individuals receive when they do not have sufficient means to live. Even though there is no incompatibility with labour income, as soon as people get a job, they earn too much to be entitled to the minimum wage. It follows that employed individuals usually do not receive a minimal wage and that even people with unemployment benefits are not entitled to this allowance. Important to notice is that childcare benefits and alimony from the ex-partner are not taken into account when determining entitlement to the minimum wage. The allowance is also increased when the beneficiary has minor (dependent) children.

Individual employment trajectories were then clustered in order to detect typical patterns of labour market participation after lone motherhood. In sequence analysis, the relationship between sequences is understood as the distance between them. It follows that prior to the clustering procedure on sequence objects, the computation of a pairwise dissimilarity matrix is needed. We used a variant of the classical

<sup>4</sup>All the analyses on sequences were performed using the software R version 3.1.2 (R Development Core Team 2014). The sequence analysis was conducted by using the R packages TraMineR (Gabadinho et al. 2011) and WeightedCluster (Studer 2013).

Optimal Matching called Hamming distance (HAM) to compute the matrix due to its higher sensitivity to small time changes (Studer and Ritschard 2014) 5 .

Individual sequences were then clustered in order to maximize within-group homogeneity and between-groups heterogeneity: to this purpose, we used the partitioning around medoids method. Medoids are representative sequences having the smallest dissimilarity to the other sequences of the cluster they belong to6 . According to the average silhouette width criterion used for detecting the number of clusters fulfilling these requisites (Kaufman and Rousseeuw 2005), an eight-cluster solution was detected as the optimal one7 .

In the following analytical step, we estimated the differential probability of following those emerging patterns according to the variables of interest. We ran a multinomial logistic model in which the dependent variable was cluster membership. The independent variables refer to both individual and household characteristics: age of the mother 1 year before the transition to lone motherhood occurred (25–30, 31–35, 36–40, or 41 or more); civil status before lone motherhood (cohabiting or married); presence of children younger than 2 years old in the household (yes or no); and number of children younger than 17 years old in the household (none, 1, 2, or 3 or more). The last two variables were measured 1 year before the transition to lone motherhood occurred. At present, the Belgian register data do not include information on the level of education. Table 12.2 shows the distribution of the independent variables considered in the analyses. Results from the multinomial regressions will be presented as average marginal effects (Wooldridge 2002), meaning that we will consider whether differentials exist in the probability of being assigned to each of the clusters according to pairwise comparisons between the categories of the independent variables and, if so, whether they are significant (Long 1997)8 .

Finally, we connected previous results with additional empirical evidence concerning household income and partnership trajectories after lone parenthood. In the first case, we built individual sequences starting 1 year before and interrupted after 5 years that account for the position in the income quintile distribution at each point in time. For the partnership trajectories concerned, we consider the survival probability function of exiting lone motherhood by entering a new partnership. These additional descriptive results helped us to shed light on possible interaction between family/employment domains and the risk of economic vulnerability.

<sup>5</sup>A number of distance measures were tested and the structure of the clusters was substantially the same (results from the sensitivity tests are available upon request). We then selected the HAM option in order to maximize the timing sensitivity according to our substantive interest. HAM distance has all substitution costs equal to one and no *indels.*

<sup>6</sup>The corresponding algorithm pursues a global optimization: in contrast, hierarchical methods usually follow a local optimization that instead could not be optimal on a global level (for details, see Gabadinho et al. 2011).

<sup>7</sup>The average silhouette width (ASW) (Kaufman and Rousseeuw 2005) measures the coherence of the assignment of each sequence to a cluster and its value is allowed to vary from −1 to +1. In our case, the ASW value for the 8-clusters solution using the HAM distance was 0.51.

<sup>8</sup>Marginal effects estimations were performed by using the SPost13 Stata commands (Long and Freese 2014).


*LM* lone motherhood. Data source: CBSS Datawarehouse 2009. Authors' own calculations

#### **Results**

In order to address our first research question concerning the variability of employment trajectories during and after lone motherhood, we started by running a cluster analysis on individual employment trajectories. The eight-cluster solution is displayed in Fig. 12.1. A general characteristic emerges from the very first sight: three clusters – namely "1. Reducers", "3. Increasers", and "5. Returners" – show a progressive shift from one prevalent state to another, while the remaining five clusters are characterized by a steady permanence in a single state. In other words, the first macro-group gathers the "movers" and the second one the "stayers". However, when we look at the "stayers" clusters, the importance of taking into account the year before lone parenthood (represented by the time point preceding the 0 value on the x-axis that represents the year of the transition to lone motherhood) emerges clearly. In fact, we see that also for the "stayers" the actual dynamic of employment trajectories implies a change from before to after lone parenthood, signalled by the fact that the prevalence of a certain arrangement (for example, full-time work for cluster "2.Full-timers") changes in correspondence to the transition to lone motherhood becoming even stronger.

**Table 12.2** Distribution of the independent variables

**Fig. 12.1** States' distribution plots for the 8-cluster solution accounting for 1 year before and 5 after the transition to lone motherhood (LM). Data source: CBSS Datawarehouse 2009. Authors' own calculations

Fig. 12.1 makes clear that there are different ways to move/stay within and outside of the labour market structure. In fact, among the movers, those who belong to cluster "1.Reducers" progressively reduce their working hours, moving from fulltime to part-time employment, while those who are assigned to clusters "3. Increasers" and "5.Returners" increase their attachment over time. It is interesting to notice that such shifts start right before the end of the second year after the transition to lone motherhood. This can be interpreted – in the case of "1.Reducers"– as the consequence of an increasing need of care within the household, which forces mothers to reduce their labour supply after a first attempt to conciliate the previous arrangement and the new family arrangement; or–for the "Increasers" and the "Returners" as a response to the increasing need of income.

All in all, those who permanently stay within the labour market ("2.Full-timers" and "3.Part-timers") represent more than 50% of the sample, while lone mothers who are steadily excluded make up 15%. In three cases – "6.Ousiders", "7. Subsidized", and "8.Marginalized" – the transition to lone motherhood seems not to alter the previous arrangement. In these cases, we might think that lone motherhood acted by weakening the probability of exiting a situation of unemployment/inactivity. This might have made those women even more dependent on external help and undermined their employability9 .

To estimate the probability of lone mothers being engaged in the different patterns of labour market participation according to their own and household characteristics,

<sup>9</sup>Distribution of individuals' characteristics within each cluster and clusters profiles can be found in Tables 12.A1 and 12.A2 in the Appendix.

we estimated a multinomial logistic regression with cluster membership as the dependent variable. The results are presented in the four plots in Fig. 12.210:

Plot (a) in Fig. 12.2 shows the probability of being assigned to each cluster according to age of the mother 1 year before the transition to lone parenthood. As mentioned above, our interest on this individual characteristic is two-fold. Firstly, young lone mothers might have experienced only short periods of paid work before the transition to lone parenthood and thus have less bargaining power to direct their strategy when the need arises to rearrange the labour supply. Secondly, both early motherhood and lone motherhood might have prevented them from catching up on their education, and low education represents a disadvantage in the labour market in terms of income and employment security. Two patterns emerge clearly from plot (a). Firstly, being a young lone mother (25–30) is associated with a significantly higher probability of being "5.Returners" or a "6.Outsiders" and to a lower probability of being "2.Full-timers" and "4.Part-timers" but also "8.Marginalized" compared to older lone mothers. The "8.Marginalized" pathway is an extremely disadvantaged one, which is more frequently followed also by women who experience lone motherhood when they are older than 41. The latter are less likely to be "2.Full-timers", but at the same time more likely to increase their labour supply over the years (cluster "3.Increasers") compared to younger women.

Cohabiting instead of having been married before lone motherhood (plot (b) in Fig. 12.2) is associated to a higher probability of being "5.Returners" or "6. Outsiders", as the probability of cohabiting itself is higher for younger women and they are more likely to belong to those clusters. In contrast, the other clusters are less frequently associated with cohabitation than marriage prior to lone parenthood. These results seem to be unexpected if we consider that cohabitation is usually associated with a higher employment rate for women compared to marriage, while the two clusters in which ex-cohabiter lone mothers prevail show a relatively weak labour market attachment, if any. However, this might be due – once again – to the relatively younger age of lone mothers in these clusters, whose age and early motherhood might have prevented them from starting their own employment career.

When we move to consider the characteristics of the household, we basically account for variations in the care-giving investment mothers might face when the household structure changes. Plot (c) in Fig. 12.2 shows the average marginal effects of the presence of children younger than 17 in the household. Two main trends emerge. When compared to having only children older than 17 in the household, the presence of one or more younger children is associated with a higher probability of being "4.Part-timers" and a progressively lower likelihood of being "2.Full-timers" as the number of children increases. Moreover, the probability of being "6. Outsiders", "7.Subsidized", and "8.Marginalized" is higher for women caring for a higher number of children.

<sup>10</sup>Tables reporting the full results of the multinomial logistic regression model are available upon request.

**Fig. 12.2** Linear predicted probabilities for the assignment to each cluster: "1.Reducers", "2.Full-timers", "3. Increasers", "4. Part timers", "5.Returners", "6.Outsiders", "7.Subsidized", "8.Marginalized". 95% confidence intervals are displayed. *LM* lone motherhood; *HH* household. Data source: CBSS Datawarehouse 2009. Authors' own calculations

In addition, we tested for the association between employment patterns and the presence of a baby aged 0–2 in the household, which is likely to strongly increase the need for care. Estimates displayed in plot (d) in Fig. 12.2 show that the presence of an infant is significantly negatively associated only with a higher probability of being "4.Part-timers". In other words, it appears that the number of children to take care of rather than the very young age of one of them either represents a stronger barrier to certain pathways or prevents certain changes and thus traps mothers in certain disadvantaged positions.

Our last analytical step aims to highlight the connection between the employment pathways after lone motherhood and household income and partnership trajectories. Fig. 12.3 shows the income trajectories experienced by women belonging to the 8 clusters. The over-plotted black lines represent the survival probability to enter new partnerships for those same women. Considering the quintile income distribution starting from 1 year before the transition to lone motherhood allows us to establish evidence that this event triggers a worsening during the first year after its occurrence. The only exception is represented by cluster "7.Subsidized", since the proportion of women in the second quintile increases against a decrease in the third and in the first. In general, not many changes affect the experience of the "stayers" income-wise, but of course this occurs to a different extent depending on the fact of being in or out of the labour market. Marginalized lone mothers (cluster 8) are those most penalized, as they have neither unemployment benefits nor a minimal wage. Their household income remains steadily in the bottom quintile, even though half of

**Fig. 12.3** Distribution across the income quintiles for the 8-cluster solution from 1 year before to 5 years after the transition to lone motherhood. Survival curves for exiting from lone parenthood and re-partnering by starting a new cohabitation are over-plotted in white. *LM* lone motherhood. Data source: CBSS Datawarehouse 2009. Authors' own calculations

them re-partner within 3 years from the transition to lone motherhood – which is a higher rate compared to the other groups of stayers, among whom only 35–40% of the women re-partner within the same time span.

Not surprisingly, a bigger number of income trajectories of full-timers than of part-timers are characterized by episodes in the top quintiles. However, when we compare these two groups to those who ended up with the same labour market arrangement (full time and part time, respectively) but started differently, we see that moving from part time to full time ("3.Increasers") led to a final income distribution characterized by a narrow top and a wider third quintile compared to that of "2.Full-timers". The same applies to the comparison between the "5.Returners" and the "4. Part-timers": the former move from unemployment/inactivity to part-time work, and they are massively distributed in the second and third quintiles. In contrast, almost 25% of the "4.Part-timers" are distributed in the top two quintiles.

This evidence further stresses the importance of seniority within the labour market in defining income profiles that then interact with needs, opportunities, and constraints lone mothers might encounter. Interestingly enough, the "1.Reducers" and the "5.Returners" are those who – on average – were more likely to re-partnered within the first 4 years after lone parenthood (55%). In the first case, the income might be more favourably distributed on the top quintiles (despite having switched from full-time to part-time employment) because of re-partnering. In fact, we might expect a general decrease in the household income due to the reduction in the labour supply, but this seems to be buffered by an additional income from the new partner. In contrast, the change in the income distribution for the "5.Returners" is more likely to be a pure "employment effect" since re-partnering time for this subgroup is more delayed.

#### **Concluding Remarks**

In this chapter we provided descriptive evidence from the Flemish case on the variability of labour market participation patterns after women experience lone motherhood. A single-country study allowed us a better understanding of processes that are strongly shaped by institutional configurations and individual outcomes over the life course (Mayer 2005). Exploiting longitudinal data from the CBSS Datawarehouse, we could follow individual trajectories in the labour market or as welfare recipients during 4 years after this family transition occurred. Our exploratory analysis aimed to account for the complexity of within-group differences by identifying typical trajectories and estimating the relative probability of experiencing them according to individual and household characteristics. We chose to focus on within-lone-mother-group differences to highlight life-course phases of high vulnerability and potential social exclusion defined by the intersection of employment and welfare for individuals who experience this family transition.

Our main results support a complex understanding of lone mothers' experience of paid work and welfare (Zagel 2013). In fact, the typology we obtain from clustering individual sequences reveals that there are many strategies lone mothers adopt when they have to cope with a new family composition. The variety of employment (and welfare) behaviours over the life course during the specific temporal window of "post-lone-parenthood" is consistent with previous findings adopting similar analytical strategies (Haux 2013; Zagel 2013; Stewart 2009). Furthermore, according to previous findings (Eamon and Wu 2011), more than 50% of the lone mothers in our sample have a strong labour market attachment, being and remaining in full- or part-time work for the whole time span considered. Another 12% either return after unemployment or increase the labour supply within the first 2 years after the transition to lone parenthood has occurred. In other words, the debate on the economic cost of lone parents actually concerns only a part of those who experience lone parenthood at some point during their life course. In our case, 15% of the lone mothers in the sample experienced long-lasting exclusion from the labour market, as they were either unemployed/inactive (cluster "6.Outsiders") or receiving the minimum wage (cluster "7.Subsidized"). However, a not negligible 10% does not receive welfare support in the absence of employment for all 4 years considered (cluster "8. Marginalized"). The persistence in such configurations reveals the presence of disadvantages in place even before lone parenthood, which we unfortunately cannot account for due to the lack of information – for example – on education or migration status in register data.

Nevertheless, we found the individual and household characteristics we considered as possible "barriers" to employment are differently associated to the patterns we identified. In this case, the within-group variety appears to be crucial in defining more vulnerable profiles: young lone mothers seem less able to engage in paid work and thus experience relatively long periods as recipients of welfare measures, while having had the time to build their careers before lone motherhood seems to keep older lone mothers remain strongly attached to employment. In other words, the age at which they have children and/or experience lone motherhood is crucial in understanding their future exclusion from the labour market. In the new social risks framework (Bonoli 2005; Vandecasteele 2011), policies should address exactly such situations in which long-term disadvantages that emerge when the timing of life transitions in certain domains matches with traditional factors of inequalities such as lower education.

Finally, in the Flemish case it is the number of children below 17 in the household rather than the presence of very small children that defines different probabilities of pursuing a certain labour market strategy. Specifically, the bigger the family size, the higher the probability of having a strong labour market attachment through full-time jobs, and the lower the likelihood of being unemployed/inactive and receiving welfare benefits.

Besides the strength represented by the use of longitudinal register data on labour market participation and by the possibility to make a fine-grained distinction between types of welfare support and employment arrangements, some limitations of this work have to be considered. First, a specific group of lone mothers falls out of our scope because the initial sample of our study comprised only women who either were married or were cohabiting before lone parenthood. Women who live at their parents' house or are lone mothers by choice remain out of sight in this analysis. It follows that we might underestimate the incidence of trajectories characterized by unemployment and welfare dependency because the frailest women (lone mothers at first birth and very young lone mothers) are not in our sample. We believe this bias is limited in this case because Belgium scores very low (0.6%) on fertility rate of women between 15 and 19 (World Bank 2015). The rate of teen pregnancies is also low and stable around 9% (De Wilde 2008). As shown elsewhere in this book, many other issues also concern lone parenthood at this young age besides labour market strategies.

This data limitation applies to another crucial issue: the Belgian register data currently do not report individuals' educational level. When analysing employment trajectories, education represents a potential confounder when estimating the probability of being more or less attached to the labour market, and this is particularly true for women (DiPrete and Buchmann 2013).

Further research should face these limitations, but we believe that our contribution adds to the recent body of evidence on lone mothers' strategies on the labour market. Together with Berrington (2014), Haux (2013), Whitworth (2013), and Zagel (2013), our single-country case study again shows the complexity and diversity of lone parenthood, more specifically among lone mothers. Survey data often limit researchers' ability to dip deep into this complexity, while registers are currently limited to a single-country approach. Future collaborative studies may focus on combining the power of registers in a comparative endeavour that aims to combine labour market strategies and income positions cross nationally and disentangle the subtle influences of welfare states on these strategies. This is important since we have shown in this chapter that a majority of lone mothers are active on the labour market. It is the small group of frail lone mothers that warrants our attention because they are most likely to struggle with poverty and perpetuate their disadvantaged position in their future life courses.

**Acknowledgements** This publication benefited from the support of the Swiss National Centre of Competence in Research *LIVES – Overcoming vulnerability: life course perspectives*, which is financed by the Swiss National Science Foundation. Emanuela Struffolino is grateful to the Swiss National Science Foundation for the financial support. The authors also thank the Belgian Crossroads Bank of Social Security and more particular Chis Brijs for their cooperation in this research.

This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).



**Appendix**


#### **References**

Abbott, A. (2001). *Time matters. On theory and method*. Chicago: University of Chicago Press.


**Emanuela Struffolino** is senior researcher at the "Demography and Inequality" research group at the WZB Berlin Social Science Center and at the chair of Microsociology at Humboldt University of Berlin. She obtained her PhD in Sociology at the University of Milano-Bicocca and worked as research fellow at the National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life course perspectives (NCCR Lives), Lausanne, Switzerland. Her research interests include life-course sociology, gender inequalities in the labor market, social stratification, and methods for longitudinal data.

**Dimitri Mortelmans** is Full Professor of Sociology at the Faculty of Political and Social Sciences of the University of Antwerp, Belgium. He teaches qualitative research methods, applied multivariate statistics and advanced topics in family sociology, life-course sociology and demography. He is head of the Centre for Longitudinal and Life Course Studies (CLLS). His research concentrates on family sociology and sociology of labour. He has published on divorce, new constituted families, gendered labour careers and work–life balance. He is also the main author of the Step in Statistics book series of which six volumes have been published (in Dutch). On qualitative methodology, he published the Handbook of Qualitative Research Methods and Qualitative Analysis with Nvivo.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Part V Well-Being and Health of Lone Parents**

## **Chapter 13 Associations Between Lone Motherhood and Depression: A Co-twin Control Study**

**Diana Dinescu, Megan Haney-Claus, Eric Turkheimer, and Robert E. Emery**

#### **Introduction**

Both in Europe and in the United States, lone mothers receive more condemnation than sympathy from the media, politicians, and even religious figures. Lone mothers are blamed for their children's bad grades, for poverty, and even for gun violence. This makes it all the more important to understand the effect that their parenthood status has on their mental health and well being. Research shows that women are more likely to be depressed than men, controlling for age, education, and employment (Atkins 2010; Collings et al. 2014), and lone mothers' risk increases three fold compared to married mothers (Cairney and Wade 2002; Collings et al. 2014). Although in recent decades policy measures have begun to be put in place with the intention to support lone mothers and alleviate their struggles, both Europe, and especially the US, have farther to go.

This chapter will blend research and policy reviews with quasi-experimental results in order to provide more insight into the relationship between lone motherhood and depression. While we will begin by reviewing research that examines lone motherhood using traditional methods, these methods do not fully explain the association between lone motherhood and mental health outcomes such as depression. Thus, we will introduce a useful approach for studying behavioral data, and one that has never been used in the study of lone motherhood: twin research. We will explain its theoretical underpinnings and its practical applications in the matter that is of interest to us: lone motherhood. We will then introduce the case study we will use to illustrate this method. Due to Sweden's generous policies aimed at lone parents,

University of Virginia, Charlottesville, VA, USA e-mail: dd7kc@virginia.edu

D. Dinescu (\*) • E. Turkheimer • R.E. Emery

M. Haney-Claus College of William and Mary, Williamsburg, VA, USA

we will use a Swedish twin sample to investigate the relationship between lone motherhood and depression levels. We will present our methods and results, and discuss our conclusions. Based on our results and the current policies in Sweden, we will discuss implications and possible future directions.

#### **Prior Research Investigating Lone Motherhood and Depression**

Research conducted by teams all over the world (Europe, Canada, the United States, Australia) suggests that lone mothers have a higher risk of depression than married mothers (Cairney et al. 2003; Colton et al. 2015; Crosier et al. 2007; Robinson et al. 2014; Sperlich et al. 2011a; Wang 2004). Women have been found to have an almost doubled risk of depression when compared to men (Atkins 2010), and lone mothers are three times more predisposed to poor health and depression than their married counterparts (Cairney et al. 2004; Colton et al. 2015). Age and race are correlates of this association, with younger women being more predisposed to depression than their older counterparts, (Brown et al. 2012; Horwitz et al. 2007), and lone mothers of color having a higher risk of depression than white mothers (Wang 2004). The number of children is an important predictor of maternal mental health, with more children corresponding with an increase in risk for depression (Horwitz et al. 2007; Sperlich et al. 2011b). Additionally, the amount of social and partner support the mother perceives to have, and chronic stress were also associated with increased risk (Cairney et al. 2003; Sperlich et al. 2011b).

Demographics cannot account fully for this association. Economic hardships were found to be more common in lone-parent families and the resulting strain was associated with an increased vulnerability to depression (Cooper et al. 2008; Crosier et al. 2007). This is explained by considering the resources that better financial status can provide, and that can enhance health and well being: better housing, better nutrition, and better access to health services (Colton et al. 2015). Employment status is an important factor for financial stability: the depression risk of unemployed women is 27% higher than for employed women, and conversely if the women are employed full time the risk is 29% lower (Zabkiewicz 2010). In fact, maximum health benefits related to labor come from employment that is full-time, stable, and long term (Evenson and Simon 2005; Wang 2004; Zabkiewicz 2010). However, the picture is more complex when considering mothers.

A full-time job leads to a higher income which is beneficial to lone mothers' mental health (Evenson and Simon 2005), and staying employed can decrease their risk of depression by about 3% a month (Zabkiewicz 2010). On the other hand, there is evidence that full-time employment may have a negative impact on health, while part time or no employment has a beneficial health effect, when adding children into the analyses (Hewitt et al. 2006). Findings on unemployment are also inconclusive: while some studies found that unemployed lone mothers had particularly poor health (Fritzell 2011), other studies, when comparing different policy regimes, found only a marginal direct effect on lone mothers' excess risk of poor health, but did describe a synergy effect between lone motherhood and unemployment (Fritzell et al. 2012). However, it is possible that these differences were due to studies being conducted in different countries with different policy measures aimed at supporting lone mothers. We consider the research on policy regime differences below.

Welfare regime theory is increasingly used in research on the associations between social determinants and health outcomes. This theory argues that population health should be better in social democratic regimes, or in our case that supportive policy measures would help alleviate lone mothers' financial strain – and through it, the risk for depression. However, research is inconclusive as to whether different policy regimes have an impact on the association between lone motherhood and mental health outcomes. Some studies find that lone mothers have significantly worse health than coupled mothers regardless of policy regimes when comparing Italy, Sweden, and Britain (Burstrom et al. 2010), or more generally, that the type of welfare regime does not have a significant impact on health disparities in the population (Brennenstuhl et al. 2012). Other studies reveal more of a struggle to meet basic needs in countries with less support, such as Russia and the US, compared to countries with more support, such as Norway (Brown 2008), and find that lone mothers are at higher risk for depressive feelings in the Bismarckian welfare regime, out of a sample of 27 European countries (Van de Velde et al. 2014). More research is needed in order to understand the complex ways in which policy regimes impact the relationship between lone motherhood and mental health outcomes.

#### **Social Selection vs Social Causation**

Early models of health emphasized biological influences, but scholars now recognize that social factors are an equally important influence on health outcomes (Carr and Springer 2010). Within these, social selection and social causation emerge as dominant alternative explanations proposed for the associations between marriage and better health outcomes (Beam et al. 2011; Emery et al. 2012; Horn et al. 2013), socioeconomic status and psychiatric disorders (Johnson et al. 1999), social class and depression (Ritsher et al. 2001), and in many other such investigations. Researchers continue to debate the extent to which social causation or social selection processes are at play in observed associations between social determinants and health outcomes. We are proposing a study of lone motherhood and depression which uses a special type of data that will help answer the question of whether selection or causation mechanisms are at play in an observer association between our variables of interest.

In the debate between social selection and social causation, the problem arises in trying to tease out correlation from causation. Saying that lone motherhood causes depression is inaccurate if all we have to go on is correlational research. There are

two possible mechanisms for processes that could explain the relationship between lone parenthood and any kind of psychological outcome: selection and causation. Selection implies that people with positive traits or who exhibit behaviors influenced by their early rearing environments or genotypes tend to select into partnered parenthood (we will use the term "partnered" instead of "married" because we want to account for mothers who choose to be in committed long term relationships without getting married), whereas individuals with more negative traits are at a higher risk of remaining single or getting divorced. Causal pathways imply that intrinsic aspects of partnered parenthood are protective to those who enter into the parenthood state, through mechanisms such as economic advantage, the normative aspect of a relationship, and increased emotional and social support (Emery et al. 2012; Turkheimer and Harden 2013).

In other words, it is possible that depression preceded lone motherhood, or contributed to it. Alternatively, there could be extraneous factors that influence both, such as race, education, socioeconomic status, or upbringing. For instance, growing up in a lone parent home may be more likely to lead someone to become both depressed *and* a lone mother. People do in fact non-randomly self-select into marital statuses based on their parental marital status, race, and their own mental well-being (Hope et al. 1999; Mastekaasa 1991). Research has also found non-random genetic selection into phenotypes (i.e. observable characteristics) such as divorce (Johnson and Wu 2001; Simon 2002). An additional concern is that of potential geneenvironment correlation. Someone's genetic makeup is correlated with their environmental experiences, creating a potential extraneous factor that could explain the correlation between, for instance, lone motherhood and depression. In the absence of random assignment, causal inferences are difficult to accomplish. Lone mothers differ from partnered mothers in many ways, some measurable and others impossible to measure. The genetic and environmental selection factors explained above may account for an observed association between, for example, lone motherhood and psychological well-being indicators. Twin research is a robust tool that allows us to account for the presence of possible selection factors and perform a quasiexperimental analysis of correlation and causation. This method has been used extensively in investigations of observed social and behavioral outcomes, from marriage and family research (Beam et al. 2011; Emery et al. 2012; Horn et al. 2013) to BMI and alcohol use (Dinescu et al. 2015; Kaprio 2015).

#### **The Utility of Twin Research in the Study of Behavioral Data**

To the extent that lone motherhood causes differences in depression levels, when we look at twin pairs we should observe this relationship within pairs (i.e. between the two twins in a pair) as well as between them (i.e. comparing one twin pair to another). Within pairs of identical (i.e. monozygotic, or MZ) twins who differ in "motherhood status" (i.e. lone vs. partnered mothers), a causal hypothesis implies that the partnered member of the pair will display different depression levels than

the lone twin (Turkheimer and Harden 2013). This might be because partnered mothers have more opportunities to take breaks from the stress of parenting, or because they get more emotional support from their partner, or simply because they get to share the burden of parenting and responsibility with someone else. An association between lone motherhood and depression within pairs of MZ twins controls for selection factors arising in either genotype or rearing environment, because the twins are genetically identical and were raised in the same home. In contrast, if the association between lone motherhood and depression is the result of non-causal genetic or environmental confounds, the association would be observed between pairs (families in which mothers tend to be partnered tend to have positive outcomes) but not within them (Turkheimer and Harden 2013). Associations within twin pairs do not conclusively prove causation, because without randomization it is ultimately not possible to control all potential confounds, but because twinships control for genetic and shared environmental confounds at the level of families, they are a step closer than non-genetic research to a causal interpretation of the results (Turkheimer and Harden 2013). We prefer the term quasi-causal to refer to an association between uncontrolled variables that survives testing with a quasiexperimental design in which genetic and shared environmental confounds are controlled.

The utility of twin research can be illustrated by considering the case study of a pair of MZ twins discordant for parenthood status, i.e., one twin is a lone mother and the other one is a partnered mother. Let's imagine that the lone mother also reports significantly higher depression levels than her co-twin. This difference could not be caused by genetic differences, because MZ twins are by definition genetically identical. Moreover, identical twins share many measured and unmeasured environmental experiences such as cultural background, religious upbringing, and family rearing environment; thus differences in depression also cannot be attributed to shared-environmental confounds. Any difference that we observe between the lone mother and her partnered co-twin in our case study must be attributable to the non-shared environmental differences between them. In our case, that is parenthood status, which is the basis on which they were selected. Generalizing from this illustration, we conclude that twin studies offer a quasi-experimental test of whether the relationship between parenthood status and depression is the result of selection or causation.

The advantage of studying twins can be made even clearer by illustrating it graphically. Figure 13.1 presents the difference in selection confounds controlled in analyses involving unrelated adults (no control for genetic or shared environmental confounds), siblings (they share 50% of their genes, on average, and grow up in the same family at different times), DZ twins (they share 50% of their genes on average, and are reared together simultaneously), and MZ twins (they share 100% of their genes and are reared together simultaneously). Selection effects are increasingly evident with each successive comparison; in other words, the difference in depression levels between lone and partnered mothers becomes smaller. The results are not always straightforward: it is possible to encounter various outcomes, based on the selection effects that might be involved. If there is evidence of genetic selection, we

**Fig. 13.1** The illustration and interpretation of differences in selection effects obtained by studying unrelated individuals, siblings, and twins

will notice a 50% decrease in effect sizes for siblings and DZ twins, and a further decrease to 0 for MZ twins (Fig. 13.1a). If the differences are accounted for by shared environmental selection, the effects will be small for siblings, and will disappear in DZ and MZ twin analyses (Fig. 13.1b). If there are no significant genetic or shared environmental selection confounds, the effect sizes should not differ between the groups (Fig. 13.1c). Lastly, a suppression effect could mask an effect that does in fact exist. What this means is that if lone motherhood did negatively impact women's depression levels, this effect would not be apparent if women who were at low risk for depression also tended to select themselves *out* of partnered parenthood. If this were the case, the effect sizes would be larger when examining siblings and twins. The magnitude of effect size differences would depend on whether the selection effects were genetically- or environmentally-based, or both (Fig. 13.1d).

We turn to a specific case study using this approach, both to illustrate the power of genetically informed data analysis, and to discuss specific findings. We pose the question: is the relationship between lone motherhood and depression accounted for by non-random selection into lone motherhood, causal consequences of this parenthood status, or both? To that end, we will use the example of Sweden, a country with generous policies aimed to support lone mothers.

#### **Sweden: Current Policy Landscape**

Even among the wealthiest countries in the world there are significant differences between family policies, and between the various levels of support lone mothers receive. The goal of the significantly helpful policies found in Sweden is quite clear: everything revolves around creating a gender-equal economy and protecting the livelihood of children. In countries with progressive family policies, the laws are meant to encourage employment rates in women (regardless of marital or family status), and foster a more equal caretaking responsibility between women and men.

Sweden has three major policies that target family economic security, gender equality, voluntary parenthood, and children's rights (Haas 1996). Sweden supplies allowances for each child, has an extremely generous parental leave policy after the birth of a child, and has a free public childcare program. These policies are universal, and all families are guaranteed certain rights (Haas 1996). These programs do not depend on the income of the child's parents, nor are they short-term or temporary. The creation of these policies is meant to encourage all parents to take on the dual roles of working and child-rearing, with the idea that the laws in place should prevent as many difficulties that might arise from that duality as possible. The laws aim for a horizontal equality among families, regardless of family structure (i.e. lone parent families headed by lone mothers or fathers), number of children, racial or socioeconomic background (Ozawa 2004).

In order to gain this horizontal equality, Sweden's laws aim to keep as many families as possible above the poverty line. This is especially crucial when talking about lone mothers. Oftentimes lone mothers must provide all care-giving and

economic support for their children. In countries without good family policies, many lone mothers are forced to only work part time in order to care for their child due to high childcare prices. When adequate government policies are not in place to offer support, lone mothers are at a greater risk for poverty (Brady 2006; Misra et al. 2007b). In Sweden, however, lone mothers have less than a 2% risk of poverty. This statistic is quite impressive when considering that lone mothers in the United States have a risk of over 35% (Misra et al. 2007a).

Sweden's success in assisting parents, lone mothers in particular, comes from a few key policies. The first of such policies is the child allowance policy. The benefit, as of March 2014, is split equally between the child's guardians if the child is under joint custody, while lone parents get the whole allowance of about \$150 USD per month until the child turns 16 (Nordic Social Insurance Portal-Family Benefits n.d.). This policy displays gender equality as well as the encouragement for women's economic independence, even when in a heterosexual cohabitation. Research has found that when governments give families this type of assistance, poverty is more easily eradicated (Christopher et al. 2002; Christopher 2002; Kenworthy 1999; Misra et al. 2007a; Smeeding 2005).

A second important policy is the parental leave policy that is in place after the birth of a child. A substantial leave policy has been shown to help all mothers, but lone mothers benefit more. In fact, a moderate to strongly generous leave brings lone mothers down to an almost equal risk for poverty as partnered mothers (Misra et al. 2012). Parents share a maximum of 480 days of parental leave, with 2 months exclusively for each parent in order to encourage fathers to spend more time at home with their child. The first 390 days are compensated based on income. If the parent has a low income or no income at all, a basic flat rate is paid. The other 90 days are paid a low flat rate. Altogether, parents are eligible to receive up to 80% of their income while on leave (Wells and Bergnehr 2014). Leave is designed to allow women to have children, but still keep their competitive and higher paid jobs. 81% of mothers with children under the age of 18 are employed (Barnombudsmannen 2010), and Sweden's wage gap is lower than the OECD average (OECD n.d.).

Swedish subsidized childcare also keeps women from having to choose between having a child and having economic stability. It has been shown that universal childcare helps to increase both women's participation in the labor market as well as women's wages (Pettit and Hook 2005). As a matter of fact, one of the greatest deterrents that prevent lone mothers from going back to work is lack of affordable childcare (Burstrom et al. 2010). One study showed that childcare provided for children of ages 0–2 was significant in helping reduce the risk of poverty in lone mothers. But in countries where private childcare is the main source of care, despite having many working mothers, the level of poverty is still quite high (Misra et al. 2012). Thus, Sweden's policy is a dual investment; by supplementing childcare, mothers do not have to cut back to a part-time job, and children start out in a stimulating, safe, and educational environment.

We have chosen Sweden as our case study due to the generosity of the policies outlined above. The supportive policy measures in this country create a more equal environment, thus possibly limiting the environmental variability that could affect our variables. A twin sample from Sweden would thus offer the potential for limited environmental variability in addition to the possibility of controlling for genetic and shared environmental confounds. Such an investigation will help us distinguish between selection and causation and will bring us a step closer to understanding the mechanisms behind the association between lone motherhood and depression. Below we describe the methods and results of our study.

#### **A Data Illustration Using the Swedish Twin Registry**

#### *Methods*

Data for the current study was drawn from the Swedish Twin Registry (STR), the largest twin registry in the world. The STR contains information from 170,000 Swedish twins born since 1886. Several birth cohorts make up the registry totaling 29,030 MZ male twin pairs, 27,372 MZ female twin pairs, and 28,686 DZ twin pairs. Data were collected through online, telephone, and mailed questionnaires, and gathered information regarding diseases with a focus on exposure during young adulthood and midlife. We analyzed data from a cohort of the Swedish Twin Registry (STR), the Swedish Twin study of Adults: Genes and Environment (STAGE). This cohort included twins born between 1959 and 1985. The information collected included history pertaining to physical and emotional health, work history, family composition, alcohol and tobacco habits, and major life events. Of the 42,582 eligible twins, 59.6% responded (N = 25,364). Data for this study came from 1316 twin pairs (738 MZ; 578 DZ), and response rates for all variables were 90–95% in our sample.

For the purposes of the current analysis, we focus only on parental status, depression score, education level, and number of children. Parental status was determined based on a measure of marital status (from a self report item we selected the answers *married* and *cohabitating* to make up the "partnered" group, and *divorced*, *separated*, and *living alone* for the "lone" group) plus a measure of number of children. Of the members of these "lone" and "partnered" groups, we kept those participants who were living with one or more biological children at the time of the survey – unfortunately the age of the children was not available to include on our selection criteria. The participants' level of depressed affect was assessed using a portion of the Center for Epidemiologic Studies Depression (CESD) scale, one of the most common screening tests for depression levels (Radloff 1977). The respondent's education was obtained from a self-report question, "What is the highest level of education that you completed/ are working on?" Descriptive statistics on depression ratings and all control variables are presented in Table 13.1.

We carried out the comparison between lone and partnered mothers in twin pairs by fitting a series of structural equation models using the robust weighted least squares (WLSMV) estimation option in the Mplus 7.11 program, used for model fitting (Muthén and Muthén 2013). WLSMV is a pairwise present estimation method,


**Table 13.1** Descriptive statistics for depression levels, age, education level, and number of children

and assumes data to be missing completely at random (Muthén and Muthén 2010). RMSEA was used as measure of model fit, with the cutoffs 0.01, 0.05, and 0.08 indicating excellent, good, and adequate fit, respectively (MacCallum et al. 1996).

We began by fitting univariate twin models, which partition the variance of an outcome into three components: additive genetic influences (A), shared environmental influences (C), and non-shared environmental influences (E). Genetic factors (A) represent the proportion of genetic material twins share that makes them similar to one another, and correlate at 1.0 for monozygotic (MZ) twins, who share 100% of their genes, and .5 for dizygotic (DZ) twins, who share on average 50% of their genes. Shared environmental factors (C) are the cumulative effect of environmental factors that make twins reared in the same family similar to one another, and correlate at 1.0 for both MZ and DZ twins, as they are assumed to fully share parts of their environment such as parental influences, socioeconomic status, or cultural background. The non-shared environmental factors (E) do not correlate between twins, as these represent experiences that are not shared, and which contribute to their unique, within-pair variance.

Next, we fit a simple phenotypic model (Fig. 13.2), in which we regressed depression score on parenthood status to obtain an estimate of the observed effect of parenthood status on depression score (parameter *bphen*). This result represents the regression of parenthood status on depression score without statistically adjusting for the mediated effect of genetic and shared environmental selection confounds,

**Fig. 13.2** Structural equation model representation of phenotypic model used in data analysis. This model is equivalent to a population-level regression

**Fig. 13.3** Structural equation model representation of quasi-causal model used in data analysis. This model controls for genetic and shared-environmental confounds

and is equivalent to a population-level association. Finally, we fit a genetically informed phenotypic regression model (Fig. 13.3), in which the effect of parenthood status on depression score was estimated holding constant any genetic and shared environmental effects common to both parenthood status and depression level (Turkheimer and Harden 2013). In this model, we simultaneously regressed depression score on parenthood status (parameter *b′phen –* see Fig. 13.3) *and* on the A and C variance components of parenthood status.

If the *b′phen* parameter remains significantly different from zero after adjusting for the A and C regressions in the model, it can be interpreted as a quasi-causal effect of parenthood status on depression level. Such a result would support the causal hypothesis, suggesting that parenthood status has an effect on depression, above and beyond genetic and shared environmental influences that might explain the association. This means that in a pair of identical twins discordant for parenthood status (i.e. one twin is married and the other one is not) we will observe significantly different depression levels between the twins. Conversely, if the uncontrolled regression parameter *bphen* is significant, but *b′phen* is substantially reduced, even nonsignificant, the selection hypothesis would be supported. This means that the apparent effect of parenthood status on depression in the phenotypic model results at least in part from uncontrolled genetic and shared environmental factors. In other


**Table 13.2** Unstandardized parameter estimates for phenotypic and quasicausal models

*Note*: Significance is denoted by asterisk: <0.05 (\*), <0.01 (\*\*), <0.001 (\*\*\*)

bphen is the full phenotypic effect

b′phen is the genetically informed phenotypic effect

bA and bC are the indirect effects of lone motherhood on depression

words, in a pair of identical twins in which one is a lone parent and one is not, we will observe the same depression levels in both twins. All the analyses used controls for number of children, education level, and age. Although using twins is a natural control for age, since twins are necessarily of the same age, we included age as an additional control variable to account for possible variations based on the age of the various pairs.

#### *Results*

Full results from the phenotypic and quasi-causal models are presented in Table 13.2. In the phenotypic model (Fig. 13.2) we regressed depression level onto a dichotomous variable representing motherhood status (i.e., partnered vs. lone), controlling for age, education level, and number of children. We found that the effect was significant (bphen = −0.15; p < 0.001), which suggests a causal link between lone motherhood and depression in the general population. Importantly, when adding the genetic and shared environmental controls and running the quasi-causal model (Fig. 13.3), our results remained significant. We found evidence of a quasi-causal effect, such that lone mothers reported higher levels of depression than their partnered counterparts (b'phen = −0.1, p < 0.001). In other words, even in pairs of MZ twins, lone mothers are significantly more depressed than partnered mothers. The quasi-causal model included the same controls as the phenotypic model: age, education level, and number of children living with mother.

Our findings suggest that there is indeed a quasi-causal link between lone motherhood and depression, which remains even after we control for number of children, education level (which we used as a proxy for socioeconomic status), and age. Since these twin comparisons control for the genetic and shared environmental confounds that could impact this association, our result is closer to a causal interpretation of the relationship between lone motherhood and depression.

These findings are consistent with the literature on lone motherhood, and confirm that lone mothers are indeed at a higher risk for depression, even when considering other important variables such as number of children or education. However, twin research adds to the previous interpretations of these findings, by being able to eliminate entire classes of measurable and unmeasurable variables from the equation. It is not enough to know that lone motherhood is quasi-causally linked to depression, it is necessary to find the mechanisms through which this link functions, and twin research bring us one step closer to finding those mechanisms. It is clear from our analyses that the cause is not only genetic or environmental selection, and it is not only related to environmental factors such as socioeconomic status or number of children.

#### **Conclusion**

Sweden boasts very generous support for lone mothers, effective public policy measures, and significant public spending in support of these measures. It is conceivable that this would lead to a decrease in the negative consequences that lone motherhood seems to have on mental health and quality of life. We employed twin research to examine the association between lone motherhood and depression in a Swedish twin dataset. Twin research has significant advantages over traditional analysis methods, and is a useful tool in attempting to approach a causal interpretation of non-experimental data.

Our results show that even in Sweden, a country with exemplary policies aimed at helping lone mothers, lone mothers still report significantly higher levels of depression than partnered mothers. This result remains even when controlling for genetic and shared-environmental confounds, such as childhood environment and other upbringing influences. We chose to study Sweden due to its supportive policies towards lone mothers. These policies provide equal support for all lone mothers, thus reducing environmental variation. Given these conditions, we would expect a high degree of similarity between depression levels of partnered and lone mothers. Instead, we observed significant differences between the two groups.

Prior data explains the mental health consequences of lone motherhood in various ways, including financial strains and unmet child care necessities. We controlled for socioeconomic status and number of children in our analyses, and our results remained strong. Moreover, the policies in Sweden address these aspects and meet lone mothers' needs to a large extent. However, our results suggest that there might be a missing link, or an unexplained relationship between lone motherhood and depression. Policies that address the financial aspect of lone motherhood, or child care, while very useful, may not *fully* solve the problem. Our results suggest that simply offering lone mothers financial or child-care support may not fully correct the negative consequences that lone motherhood has on mental health. Lone mothers may need more social and psychological support in order to bridge the mental health gap, and increase their quality of life.

Despite the strengths of our chosen method of analysis, the fact that our data is cross-sectional limits our ability to draw true causal conclusions. While the genetically informed methodology we are using is the best way to draw quasi-causal conclusions based on the type of data we have available, it is conceivable that a variable that we did not control for, and which was not part of the genetic or shared environmental variance that we were able to account for in our models, might be impacting both parenthood status and depression. Future studies should seek to replicate our findings using a longitudinal design. Our data also did not allow for discussion of race or cultural differences. Thus, our findings generalize only to populations similar to our sample, which is primarily Caucasian and representative of a Northern European country. More exploration is needed before our conclusions can be generalized to other cultures.

Our study is the first investigation into lone motherhood using a genetically informed approach. Future research possibilities include exploring twin datasets in other countries, investigating the relationship between lone motherhood and other psychological indicators such as life satisfaction, self esteem, and self-efficacy, and adding moderators to analyses. Moreover, based on our results we suggest public policies offering more targeted mental health support for lone mothers, and more individualized support dependent on their life circumstances. Lone mothers seem to be missing more than the financial contribution that a partner would make, and thus any policies will hopefully broaden their scope to include other life aspects that lone mothers might be lacking, such as psychological support.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **References**


**Diana Dinescu** received her Ph.D. in Clinical Psychology from the University of Virginia (Charlottesville, VA, United States). Her research investigates the effect of marriage and family relationships on health outcomes using genetically informed methodology.

**Megan Haney-Claus** graduated from the college of William and Mary in 2016 with a double bachelor's degree in Marketing and Psychology. She currently works as Media Specialist for the marketing agency Merkle.

**Eric Turkheimer** is the Hugh Scott Hamilton Professor of Psychology at the University of Virginia. His current research includes detection of gene-environment interactions in twin studies of intelligence, development of statistical methods for the analysis of children of twins, and use of twins as controls in quasi-experimental designs to study the effects of parenting on adolescents' outcomes. In 2009 he has been awarded the James Shields Memorial Award for his research in Behavioral Genetics.

**Robert E. Emery** is Professor of Psychology and Director of the Center for Children, Families, and the Law at the University of Virginia. He has authored over 150 articles and several book on family relationships and children's mental health, including parental conflict, divorce, child custody, family violence and associated legal and policy issues.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Chapter 14 The Selective Nature of Lone Parenthood: The Case of Ireland**

**Carmel Hannan**

#### **Introduction**

Many children from broken homes are born "losers" and so deprived of love that they grow up to be dysfunctional adults, according to a Catholic bishop. (Irish Independent, 16/8/2011).

Lone parents often receive poor media coverage and examples such as the above quote are common. In 2011, the British Press blamed poor parenting for a series of riots which took part in London and around the UK. Much media coverage was given to the moral collapse of British society and the breakdown of the family;

Mr Cameron said that family breakdown and poor parenting had played significant role. "In too many cases, the parents of these children – if they are still around – don't care where their children are or who they are with, let alone what they are doing," he said. (Telegraph, 11/8/2011).

In the Irish case, a prominent academic called for DNA testing to be used to track runaway dads who dodge financial responsibilities for their offspring, in order to tackle Ireland's worsening economic situation. Accordingly, Dr. Ed Walsh noted that;

whether a community is wealthy or poor, where you have a predominance of lone parents, you have a predominance of crime (The Irish Examiner, 20/1/2011).

Yet, what is the empirical evidence to support such claims made about loneparenting? International empirical evidence suggests that children who grow up living with both biological parents fare better than children who don't (Acock and Demo 1994; Amato 2001; Amato and Booth 1997; Cookston 1999; Flewelling and Bauman 1990; Frisco, Muller, and Frank 2007; Mayer 1997; McKeown et al. 1997; Patten et al. 1997). A meta-analysis of 92 studies conducted during the 1980s

C. Hannan (\*)

Department of Sociology, University of Limerick, Limerick, Ireland e-mail: Carmel.Hannan@ul.ie

L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_14

(Amato and Keith 1991) and of 67 studies conducted during the 1990s (Amato 2001), for example, compared the wellbeing of children from divorced and twoparent families and found that children from divorced families had significantly lower scores on a range of outcomes including educational achievement, behaviour, psychological adjustment, self-concept, social competence and long-term health.

The empirical evidence on the differences in adolescent and child outcomes across family types may however result in a large part, or entirely, from selection bias. It is therefore possible that marital dissolution is rather inconsequential. In the Irish case, while divorce rates remain low, there are significant numbers of children growing up in never-married one-parent families. The selection view of loneparenthood maintains that unmarried childbearing does not necessarily result in negative outcomes for children. In the Irish case, the majority of unmarried parents are mothers who tend to be younger and come from impoverished backgrounds when compared to the married populations. The adverse consequences of unmarried motherhood may therefore be an artefact of the pre-existing socio-economic disadvantages of these mothers. This chapter presents the empirical evidence on the association between family structure and children's wellbeing for Ireland.

#### **The Irish Case**

Like most European countries, the number of children in Ireland growing up in oneparent families continues to increase (see Fig. 14.1). Between the 1986 and 2011 Census of Population in Ireland, the number of one-parent families almost tripled so that by 2011, 24.8% of all families were one-parent families. According to Eurostat, Ireland and Latvia had the highest percentage of children living with a lone parent in the EU (OECD 2007). The growth in lone-parenthood was driven by a dramatic rise in the proportion of births occurring outside of marriage; from 5% in 1980 to 32% in 2000 (Lunn, Fahey, and Hannan 2009). Marital breakdown has also risen in Ireland although it appears that the Irish are still adverse to divorce (Fig. 14.1). Between the 1986 and 2006 Census, the total number of people whose marriages had broken down increased five-fold, from 40,000 to just fewer than 200,000 although the rate is still low by international comparisons (Lunn and Fahey 2011).

This aversion to divorce and the continuation of a late age of marriage makes Ireland an interesting case study. It points to a selection process which means that marriage is not entered into lightly. Higher rates of entry into marriage among the professional classes were a prevalent feature of Irish demography throughout the twentieth Century (Connolly 2015). On average, those who marry and then have children are older, more educated and more likely to have professional careers when compared to those who do not have children within marriage. Despite globalisation, secularisation and a rise in individualisation, the decision to marry is still constrained by economic considerations related to one's social class. It appears that love is less "liquid" in Ireland(Bauman 2003).

**Fig. 14.1** The Changing Nature of Family Formations in Ireland 1971–2011 (Population aged 16–64) (Source**:** Minnesota Population Centre IPUMS; CSO Ireland 10% random census samples)

In contrast, childbearing is more fluid where the dominant route into loneparenthood is outside of marriage (Fig. 14.1).

#### **Debating the Role of Parents**

Following from this argument, it is not family structure per se which influences child development but the characteristics of the parents. Marriage is therefore not necessarily good for child development or indeed for society. Nor is the rise in parenting guides, a product of Beck's risk society and the need for expert advice, necessarily going to improve child development. According to the selection argument, the key is the social, economic and cultural resources of the parent or parents, rather than their marital status/living arrangements.

Such an approach contrasts sharply with Brad Wilcox's *National Marriage Project* in the US. In a series of articles which have received much press coverage, Wilcox and colleagues argue that marriage matters for child development, parental income and health, societal wellbeing and a state's economic prosperity. The overall conclusion is that marriage deserves special care and should be promoted, a point made by groups opposing the Marriage Equality Bill in Ireland which came into legislation in September 2015.

According to Putnam, the class divide in marriage is widening as the traditional family thrives among the educated elites (Putnam 2015). In his book, Putnam details the wide range of factors which differentiate the parental resources found in the homes of one-parent and two-parent families. Current income is often used as an indicator of these resources yet income differences across family types are a consequence of antecedent factors. Demographic characteristics such as nationality and age, and socio-economic status, like educational attainment and socio-economic background (social class) are all important factors influencing family income, child development and the propensity to marry. Ideally, a study of marriage would need to include all these factors and their interplay in the analysis of the effects of marriage at the individual or societal level.

#### **Data Source**

This research draws on the first wave of the *Growing Up in Ireland* (GUI) child cohort study, a large-scale survey of nine-year-old children (N = 8568) sampled within primary schools in Ireland in 2008/2009. The GUI study is an extremely rich data source, incorporating school principal, teacher, parent and child questionnaires as well as time diaries, and some qualitative data collection. At the time of writing, only one wave of the data was available for analysis.

As noted by Lunn and Fahey (2011), while there is diversity in family structures, there remain a small number of dominant family types that account for the large majority of families in Ireland as a whole. Many of the less traditional family types are not really that common in Ireland and this was reflected in the GUI data. While there are good policy reasons for treating one-parent families as a single group in relation to financial support, in this analysis never-married one-parent families were distinguished from the separated, divorced and widowed parents who are not living with a partner.

In terms of two-parent families, cohabiting and married parents living together with their children were distinguished as separate groups. Step-families are however included with this cohabiting group and lone fathers are not separately analysed given the small numbers involved (see Table 14.1). In general there is an under-representation of lone fathers in the *Growing up in Ireland* data.

#### **The Importance of Selection**

Before discussing the effects of family type on child development, it is important to detail the important role of selection effects. In common with Putnam (2015) data from the child cohort study (GUI) reveal that marriage is more common among older, more educated, and more religious mothers (see Table 14.2).

In wave one of the GUI study, the average age of a married mother in a twoparent family was 40 compared to 34 in a never-married one-parent family. Almost 19% of married mothers had a degree or higher-level educational qualification compared to 12% of unmarried mothers (Table 14.2). Mothers in previously-married


*Source:* Weighted Data GUI Child Cohort, wave 1 Note:

Children in the care of grandparents, foster parents or other relatives are included here based on their current marital status and living arrangements

Information is also included where the primary caregiver was male. The number of lone father families in the sample is too small (N = 45) to examine separately

one-parent families had much in common with mothers in married two-parent families, given this selection into marriage. Separated, divorced or widowed lone-parents (the post married group) had an average age of 40 and 16% had a degree or higherlevel qualification. Never-married mothers tend to have children at a relatively young age, compared to all other women in this child cohort study. More than half of all unmarried mothers (58.5%) were less than 25 years old when they had their child compared to only 13% of married mothers (Table 14.2).

The nature of family structure re-affirms the importance of taking account of selection effects when it comes to marriage. Of the married majority (76%) of parents in the sample, only 4% were cohabiting outside of marriage with the father of their child at childbirth, as most (84%) were married when they had their child. Cohabitation rates were higher at childbirth among those who are now separated or divorced with about 10% of those who are separated or divorced cohabiting at childbirth. Of never-married lone-mothers (9.6% of the sample), almost 1 in 4 were cohabiting with the father of the child at the time of birth.

The low divorce rate in Ireland is possibly reinforcing this selection into marriage. Lunn and Fahey (2011) suggest that cohabiting unions (6.4% in the sample) and non-marital child-bearing have served to select less stable relationships out of marriage and concentrate union instability into these non-marital family situations. Marital separation is, however, much more common than divorce, so that that the overall levels of marital instability in Irish families as a whole are much closer to a European mea**n** (Fahey 2012)**.**


**Table 14.2** Differences in the characteristics of the primary caregiver (usually the mother) across family types

*Source*: weighed data *Growing up in Ireland*, child cohort, wave 1

a Most of these men are widowed

b In the unweighted data, this figures stands at 3.78%

c Current age was included in the models and not age at first birth since this could not be calculated in families where information was missing for children who had left the household

Table 14.2 includes the full range of factors which have been found to differ across families, including the height of mothers as it reflects social class background. The next section moves away from socio-economic differences in the background of mothers prior to the birth of the study child and focuses on behaviour during birth and upon the birth of the child which differs across family settings.

#### **Confounding Factors**

Selection is present when the factors that select individuals into marriage also influence child development. These factors are referred to as confounding factors; variables that affect both the probability of selection into a group and the outcome of interest. The GUI study allowed for the analysis of a range of confounding factors, including those factors detailed in Table 14.2.

In addition, the behaviour of mothers during pregnancy has been found to vary across socio-economic groups and to influence the development of the child, both in the long-term and short-term. Table 14.3 lists a number of other confounding factors taken into account here. Williams et al. (2010), for example, found that smoking during pregnancy is associated with lower educational levels among mothers and is related to a range of problems among children. Most mothers in the *Growing up in Ireland* study, regardless of family type, did not report to drinking any alcoholic beverages while pregnant, whereas smoking during pregnancy has a clear gradient with 13% of married mothers smoking while pregnant compared to twice that number of cohabiting or previously-married mothers and almost a third of all nevermarried lone-mothers admitted to smoking during pregnancy (see Table 14.3).

A history of long-term illness, including depression among mothers, puts an increasing strain on family life and has been found to influence rates of marital breakdown as well as child development. Rates of chronic illness prior to the birth of the study child were not randomly distributed in our data with 1 in 10 nevermarried mothers suffering from a chronic illness *prior* to childbirth, compared to 1 in 20 married mothers. A more detailed breakdown reveals that mothers who are now separated from their husbands and rearing children alone reported the highest rate of illness prior to the birth of the child (at 21%)**.**


**Table 14.3** Differences in the ex-ante characteristics of mothers across family types

Source: weighed data GUI child cohort, wave 1

A series of other confounding factors are highlighted in Table 14.4 which show the importance of controlling for confounding factors when trying to assess the impact of family structure on child health. Low birth weight is more common among never-married lone-mother (8% of children weighted less than 2.5 kg), compared to 5% born to a married mother (Table 14.4). This is related to the fact that never-married lone-mothers are more likely to have their child prematurely (2.9% compared to the sample average of 1.8%).

Elective caesareans were more common among married mothers with almost 1 in 10 opting for a caesarean, compared to 3.5% of unmarried lone-mothers. The mode of delivery at birth has the potential to influence the long-term development of the child, for example, Huh et al. (2012) found a link between the risk of obesity and caesarean section. Rates of breastfeeding also vary across family types. The lowest propensity to breastfeed was found among never-married cohabiting parents and never-married one-parent families (only one in three breastfed) compared to almost half (47%) of married mothers and 44% of previously-married lone-mothers who breastfed their child (table 14.4).

The differences reported in Tables 14.2, 14.3 and 14.4 points to the role of confounding factors and the importance of taking them into account when measuring differences in child outcomes across family types. There are many unobservable differences not reported here, as well as important missing confounding variables such as social class, but bearing these in mind, it is important to assess the size of


**Table 14.4** Differences in the characteristics of the study child across family types

Source: weighed data GUI child cohort, wave 1

the differential in a range of child outcomes controlling for the confounding factors discussed here.

#### **School Clustering**

In addition to the above selection effects into non-marital childbearing, lonemothers tend to be spatially concentrated in areas of disadvantage. Family setting influences child development through socio-economic resources and these resources influence the schools and neighbourhoods in which children reside. Rather than seek to explain underachievement among children from never-married and postmarried families by focusing on the living arrangement of the mother, or the individual characteristics of the mother alone, Williams (1989) argued that the impact of the social structure must also be taken into account. All the models presented in this chapter take account of the clustering of children within schools.

#### **The Propensity Score Matching Method**

Propensity score matching techniques were employed to adjust for selection bias, controlling for the clustering of children within schools, in order to assess the effect of family structure on numerous outcomes in young children pertaining to their health, psychological wellbeing and educational development.

Propensity score matching methods use an estimate of the counterfactual group to adjust for selection bias (Rosenbaum and Rubin 1983). In order to reduce the bias of confounding factors, we need to know the answer to the counterfactual question, such as what level of an outcome would a child have gained had their parents married? By definition, the counterfactual cannot be empirically observed. However, it is possible to estimate the counterfactual by matching cases that are similar on confounding factors but differ on the focal independent variable (i.e. family type).

Cases are matched on multiple confounding factors, usually between two groups: a treatment group and a control group. One could imagine that two children are matched on the same pre-existing characteristics, one of whom is living in a oneparent family and the other in a two-parent family. Matching can, in principle, be done on a range of variables, but the more variables available, the more difficult it becomes to find a matched child. Instead, matching is carried out on the propensity score, which reflects the probability of receiving treatment assignment.

The predicted probabilities of receiving the treatment were calculated from a logit model which served to match the treatment and control groups based on preexisting observed covariates i.e. the confounding variables. Then, using the propensity scores, a sample of treatment groups and their matched cases was generated. This normally consists of two groups, for instance, unmarried mothers and their matched cases in the married group. The current study extended this common

technique to match multiple treatment groups: unmarried one-parent families, previously-married one-parent families and cohabitating families were matched to a control group of married two-parent families (groups are defined in Table 14.1).

Three issues arise in implementing propensity score matching which affect the matching process. First, matching with replacement was employed as it minimises the propensity score distance between the matched comparison units and the treatment unit. This is beneficial in terms of bias reduction. Secondly, the comparison units were chosen based on the nearest-neighbour method, which selects the *m* comparison units whose propensity scores are closest to the treated unit in question. Nearest neighbour matching faces the risk of bad matches if the closest neighbour is far away. This was avoided by imposing a tolerance level on the maximum propensity score distance (calliper).

The performance of different matching estimators was assessed prior to the settling on nearest neighbour (50), calliper (0.01). The fundamental assumption for the validity of matching is that, when observable characteristics are balanced between the family types, the groups are balanced with respect to all the characteristics relevant to the child outcomes. This assumption, however, relies on having a large number of available pre-intervention characteristics. A concern therefore was the success of the matching process given the limited list of available confounding variables.

The existence of a substantial overlap between the characteristics of the treated and control groups (the issue of common support) was assessed. To satisfy the requirement of common support, observations outside the maximum propensity score for our treatment groups were dropped (treated: off support ranged from *N*: 3 to 14 or a maximum of 3.3% of cases were dropped for the never-married oneparent group).

The second issue assessed was the percentage of bias reduction between the treated groups and the control group in the matched sample which was found to be, in general satisfactory (a range of 60–98% bias reduction) except for the case of the cohabiting group of parents where the matching process was not ideal. Caution is therefore advised in relation to this smaller group of families.

#### **Findings**

The results are split into three substantive areas; measuring the influence of growing up in a one-parent family on a child's educational development, before and after matching. Secondly, measuring the effects of family structure on a child's physical health and finally, on their emotional wellbeing. The results for cohabitation will not be discussed here given the limitation discussed in the previous section in relation to the matching process.

#### *Educational Development*

Figure 14.2 plots the size and significance of the negative effects of growing up in a never-married one-parent family in relation to the educational performance of children at age 9. The graph shows the average difference, that is, the point estimates in relation to three measures of educational performance in the never-married lone parent group compared to the married control group, with the lines reflecting 95 and 99% confidence intervals.

The first point to note is that children from never-married lone-parent families scored on average 10.6 percentage points less on a math test, 7.8% less on a reading test and missed school on average 2 days more than their married counterparts (weighted raw differences plotted in Fig. 14.2).1 Taking account of selection bias, reduced the size of these differentials so that children living in never-married oneparent families were scoring 5.2% less in the math test, 3.03% less in the reading test and missing 1.5 days more at school after propensity score matching. In other words, the maths differential was reduced by half and the reading differential was 60% smaller in the matched data (as plotted in the "PS-matched" line in Fig. 14.2).

**Fig. 14.2** Estimated educational differences between children in never-married one-parent families compared to their married counterparts (matched data) (*Note:* Confidence intervals calculated from robust standard errors for the matched sample)

<sup>1</sup>The Drumcondra academic performance tests are designed for pupils in Irish primary schools and were administered to children in a school setting. Teachers were asked to record the number of days each child had missed school since the beginning of the current school year. This was used as a measure of absenteeism.

Selection effects therefore account for much of the differences in the study child's educational outcomes .

The clustering of children within schools/neighbourhood also plays a role. Figure 14.2 plots the effect sizes from using multi-level modelling techniques which take account of the clustering of children within schools (the line labelled "Clustered raw" in Fig. 14.2).This modelling takes account of the fact that children within the same school will perform similarly on the tests administered in the study and have similar rates of absenteeism. Taking account of the school clustering alone accounts for 37% of the difference in math scores, 34% of the difference in reading scores and 36% of the difference in school attendance rates between the never married oneparent families and their married counterparts (Fig. 14.2).

We will not turn to the effects of growing up in a previously-married parental setting on the educational development of children. Children in previously married one-parent families (mostly as a product of marital breakdown rather than divorce or widowhood) also appeared to have substantial differences in educational outcomes at age 9 (Fig. 14.3). These 9-year olds scored on average 6 percentage points less on the maths and reading tests, and missed school 2 days more when compared to their married counterparts. Once selection bias is taken into account the size of the differential is reduced to 3.4 percentage points for maths, 1.6 percentage points for reading and 1.5 days absent. As before, it points to the important role of selection particularly in relation to reading scores (Fig. 14.3). Maths performance is a different story with high between-school differences in the amount of time dedicated to maths.

Figure 14.3 shows again that school clustering is important when it comes to educational performance. Focusing on reading test scores, selection bias on observed pre-existing characteristics accounts for three-quarters of the difference in reading scores between children in previously-married one-parent families compared to their married counterparts. School clustering alone accounts for about 40% of the initial raw difference (Fig. 14.4). In all cases, the propensity score matching and school effects estimates are smaller than the initial weighted raw or raw difference (labelled "Raw difference" in figs 14.2 and 14.3) across families.

#### *Physical Health*

A number of child outcomes were analysed in order to assess differences in the physical health of 9 year olds growing up in differing family types. In the GUI study, the primary caregiver was asked about the number of nights which the nineyear-old had spent in hospital as an in-patient over his/her lifetime, excluding neonatal care. They were also asked to indicate how often the Study Child visits the dentist. The Child's height and weight were measured by the interviewer in the course of the interview and this was used to construct BMI. Children's dietary intake was also assessed in the study via parental recall of the Study Child's eating habits in the preceding 24-hour period using a 20-item semi-quantitative food

**Fig. 14.3** Estimated educational differences between children in previously-married one-parent families compared to their married counterparts (matched data) (*Note:* Confidence intervals calculated from robust standard errors for the matched sample)

frequency questionnaire. One aspect of this food diary is plotted in the graphs below, the amount of fruit consumed.

The analysis revealed that children from never-married lone-parent families were more likely to have spent time in hospital as an in-patient and to have higher BMI when compared to children from married two-parent families (Fig. 14.4). The increased likelihood of hospital stays and higher BMI was however related to the deprived socio-economic backgrounds of children from these families and not to lone-parenthood per se (the matched differences reported in Fig. 14.4 are close to zero and not statistically significant). There was no evidence of difference in rates of dental visits or the amount of fruit consumed in the diet between children living with lone versus married parents.

The physical health of children living in previously married one-parent families and their married counterparts differed only slightly (Fig. 14.5). Compared to educational outcomes, the differences in terms of size (effect size) were small in relation to a child's health as measured in the outcomes plotted in Figs. 14.4 and 14.5.

**Fig. 14.4** Estimated physical health differences between children living in never-married oneparent families compared to their married counterparts (*Note:* Confidence intervals calculated from robust standard errors for the matched sample)

**Fig. 14.5** Estimated physical health differences between children living in previously married one-parent compared with married counterparts (*Note:* Confidence intervals calculated from robust standard errors for the matched sample)

#### **Psychological Wellbeing**

The GUI study collected information from the children themselves on how they viewed themselves. The children were asked to complete a detailed set of 35 questions known as the Piers- Harris Self-Concept Scale (Piers 1984). It gathered information about how children perceive themselves across the six domains of:


Significant differences were found in relation to how children viewed themselves. In terms of weighted raw differences, children from lone-parents families scored substantially lower on the total overall score (Figs. 14.6 and 14.7) indicating more negative self-concept when compared to children living with married parents. These differences are greatly reduced once selection bias is taken into consideration as children from lone-parent families scored lower on the total score in the main because of the differential selection into marriage compared to all other family states. Selection effects account for between 70 and 75% of the initial weighted differences in total Piers-Harris scores.

Analysis of the subscales reveals some interesting differences from the child's perspective. Children from never-married lone-parent families were significantly more likely to report negative attitudes across all subscales, except for physical appearance. Children in never-married one-parent families had a lower average score (0.39 point lower post matching) than their married counterparts on the "Freedom from Anxiety" subscale, indicating that they more frequently reported feelings of anxiety than children with married parents, despite taking account of selection effects ("FrAnx" in Fig. 14.6). However, the size of the effects is greatly reduced in the matched data.

Children who had previously lived with their married parents but now live in a one-parent family had a lower average score (0.20 point lower post matching) than their married counterparts on the Happiness/Satisfaction subscale, indicating they reported pre- and post-matching more negative attitudes about their happiness ("Happy" in Fig. 14.7). Yet in terms of all the subscales, the differential is close to zero post –matching (see Fig. 14.7). It appears therefore that the role of selection bias in understanding differences in children's self-concept is powerful.

**Fig. 14.6** Estimated differences in Pier Harris Self-Concept between children living in nevermarried one-parent families compared to their married counterparts (*Note:* Confidence intervals calculated from robust standard errors for the matched sample)

**Fig. 14.7** Estimated differences in Pier Harris Self-Concept between children living in previouslymarried one-parent families compared to their married counterparts (*Note:* Confidence intervals calculated from robust standard errors for the matched sample)

#### **Discussion**

There are clear systematic differences in the wellbeing of children across family types in Ireland. Much of the adverse consequences of childbearing outside of marriage pertaining to the child's health, self-concept and educational development appear however to be an artefact of pre-existing socioeconomic disadvantages. Compared to married mothers, lone-mothers are on average younger and they are more likely come from impoverished backgrounds (as indicated in their recollection of difficulties to make ends meet reported in Table 14.2), with lower levels of educational attainment on average when compared to marriage mothers. The matching results reported here suggest that when faced with similarly adverse conditions when growing up, children from one-parent families fare similarly in most regards to children from married families. However, some significant and modest differences do remain especially in terms of math scores and school attendance among children from never-married one-parent families.

The effect size varies across the range of child outcomes discussed here and the role of selection bias differs, explaining less of the differences in maths scores. The matching exercise had several limitations related to a lack of longitudinal data employing only one wave of data. The matching approach adopted here addresses selection effects driven by differences in observable characteristics between mothers but there remain important unobservable differences that are not accounted for especially in relation to pre-existing factors that were not recorded in the GUI study. The analysis therefore has only partially addressed the 'selection' issue by examining the size of the gap in outcomes between children growing up in lone-parent families and those in married families after controlling for observable factors. The role of unobservable factors is not clear. An individual's decision to marry or not, and have children, will depend on many factors such as group norms, perhaps personality characteristics and availability of partners. None of these factors were taken into consideration here.

Yet, despite these limitations, the matching models have in the main substantially reduced the size of the differential in child outcomes scores. The evidence presented here strongly suggest that selection effects matter and thus point to concern with policy advise which would suggest that promoting marriage will reduce child disadvantage. The final section for this chapter will now discuss this policy context.

#### **Families, Policy and Research**

Our understanding of the social structure of inequality and the regeneration of inequality across families is growing giving the increased use of a life-course perspective as it relates to families. The availability of longitudinal data and methodological advancements has helped us to unravel the mechanisms by which families matter. Much work remains to be done especially in relation to translating these findings into a policy context.

The finding that selection matters is not unique to the Irish case (see Dohoon 2010 for a US analysis or Goodman and Graves 2010 for the UK). International research has tended to agree that children are at increased risk of adverse outcomes in one-parent families and that negative outcomes can persist into adulthood, but once account is taken for confounding factors the difference between children from one and two-parent families is a small one. While the antecedents, processes and sequelae of family breakdown vary across countries, its impact on children appears fairly consistent. In the Irish case, very few one-parent families are the result of divorce and few children are growing up in step-families, yet the findings point to an increased risk of educational under-performance compounded for children in one-parent families while the protective factors are reinforced for children in twoparent married families.

The evidence therefore indicates that much of the association between loneparenthood and negative child outcomes is related to selection bias. The apparent benefits to marriage may not be all they seem. Rather, it is the interplay of a number of risk and protective factors at the family, school and neighbourhood level that serve to influence child development in children. As highlighted in this book, family structure is a process and not just a single event and given the lack of a longitudinal focus of this chapter, the dynamic nature of change in family relationships is missing. However, it appears that the benefits of marriage are overstated as they relate to child wellbeing. Parents who are married differ from those who never-married in very substantial ways, particularly relating to educational and socio-economic status of lone-mothers. The differential selection into marriage is therefore a key factor which policy-makers must consider. The results presented here strongly indicate that encouraging parents to formally marry will do little to improve child wellbeing rather policy should focus on mother's education and socio-economic status.

In the Irish case, interventions to help children and parents of one-parent families have mostly involved income supports. According to the OECD (2006), Ireland had the third highest public spending on family benefits as a percent of 2011 GDP. Cash benefits to families were highest in Sweden and then Ireland. Some argue that the reliance on income supports has led to welfare dependence amongst the range of lone-mother families (lone fathers are more likely to be employed). Recent initiatives, in particularly a labour market activations policy, aim to have lone-mothers working when the child reaches the age of 7. In other words, the welfare reform is forcing lone-mothers (and not all mothers) into the labour force; despite the evidence showing that on average lone mothers have lower educational levels and poorer employment prospects as well as a lack of formal childcare for all mothers in the Irish context (Murphy 2014). The findings presented here point to the importance of back-to -educational supports for lone mothers, many of whom have their child at a relatively young age.

Another approach is to direct supports to the children through school services. The current roll-out of an *Area Based Childcare* programme is an initiative aimed at providing early childhood interventions at a multi-agency level. Such programmes

acknowledge that a school intervention, such as a homework club, can provide some benefits but that interventions and approaches within disadvantaged areas should be integrated with mainstream services such as health, education and family supports. In other words, the public sector is seen to take on more of a role in child education, childcare and other labour that now falls into the remit of the lone-mother. In an era of increasing concern about income inequality and growing rates of child poverty, the implications of family research for policy formation needs increased visibility.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **References**


**Dr. Carmel Hannan** is a lecturer in sociology at the University of Limerick. She has previously worked at the ESRI, in Dublin, ISER at the University of Essex and at the University of Oxford. Her research has covered a diverse range of topics including educational inequalities, unemployment and friendship patterns, policy evaluation and more recently, the effects of growing up in a one-parent family on child wellbeing. She has led a number of funded projects in this area, culminating in publications such as *Growing up in a One-Parent Family: The influence of family formation on child outcomes* (2013) and "Family Structure and Child Outcomes: Evidence for Ireland (ESR 2014). Her current research focuses on family inequality over the life-course.

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Chapter 15 Changes in Lone Mothers' Health: A Longitudinal Analysis**

**Mine Kühn**

#### **Introduction**

In all industrialized countries the number of lone parent households has increased in the last decades. This family structure is particularly interesting in international research due to the special economic and social situation of lone mothers. Since the 1990s, there is a well-established literature for United Kingdom, USA, Canada, and also for the Scandinavian countries. In contrast, research in Germany has focused comparably late on lone mothers, investigating their economic disadvantages. This is somehow surprising as the structure of German families has changed substantially over the last three decades. Census information for Germany suggests a growing rate of lone parent families. In 2012, almost 20% of all families living with underage children (in Germany children are underage until they are 18 years old) are headed by lone parents (Statistisches Bundesamt 2013, p. 104). There is an increase of lone parent households from 1.28 million in 1996 to 1.6 million in 20121 (Statistisches Bundesamt 2010, p.7).

Not only the increasing rate of lone mothers in Germany is a concern, but also their increasing economic disadvantage and the loss of wealth, especially after marriage dissolution. Studies analyzing the labor market participation of lone mothers argue that time allocation is apparently more challenging for lone mothers because they cannot rely on intra-household division of labor (Hancioglu and Hartmann 2013). The percentage of lone mothers working full-time is higher when compared to mothers with partners, but the majority of lone mothers in Germany are unable to work or have a part-time employment because of a lack of affordable childcare (Ott et al. 2011). Lone mothers in Germany are frequently at higher risk of unemployment

M. Kühn (\*)

© The Author(s) 2018 323 L. Bernardi, D. Mortelmans (eds.), *Lone Parenthood in the Life Course*, Life Course Research and Social Policies 8, DOI 10.1007/978-3-319-63295-7\_15

<sup>1</sup>German Census is collecting lone parents since 1996, therefore there is no existing previous Census data on lone parent households in Germany.

Max Planck Institute for Demographic Research, Rostock, Germany e-mail: kuehn@demogr.mpg.de

and rely on welfare above the country's average (Hancioglu and Hartmann 2013; Lietzmann 2009). These empirical findings are congruent with international research results (e.g. Francesconi and Van der Klaauw 2007; Vandecasteele 2011).

At the same time, it seems that economic and social disadvantages affect lone mother's health negatively. As a result, there has been an increasing interest in studying health consequences associated with lone motherhood. Results from interdisciplinary studies suggest that lone mothers suffer disproportionally higher rates of physical and mental illness as compared to their partnered counterparts (e.g. Butterworth 2004; Crosier et al. 2007; Wang 2004). To gain deeper insights of health inequalities in Germany, analyzing potential health consequences for lone mothers is of special interest for research and for German social policy.

Current research on the health of lone mothers explores the differences in health outcomes between lone and partnered mothers. However, most of these comparisons are based on cross-sectional data and therefore are not able to consider the dynamics of the life course of a lone mother and to identify causes and effects of health statuses. Moreover, lone mothers in cross-sectional data are often described as a homogeneous group without considering that they may face a different duration of lone mother episodes.

To our knowledge this is the only research analyzing lone mothers' health in Germany using longitudinal data. This chapter aims to overcome the described weaknesses of identifying effects of being in lone parenthood. Therefore, we use panel data to explore individual changes of lone mothers' health. Further, we examine the health consequences of becoming a lone mother focusing on the questions: (1) Does transition into lone motherhood has negative effects on health? (2) Which determinants affect women's health satisfaction and well-being within the transition into lone motherhood?

This chapter is structured as follows. Section "Literature review" reviews the literature relevant to this topic. Section "Data & methods" describes the data and our methodological approach as well as the measured health outcomes and potential determinants of lone mother's health. Section "Results" presents descriptive findings and the results of multivariate analysis. Section "Concluding remarks" concludes.

#### **Literature Review**

Previous studies in this field concentrate on the effects of being married versus being non-married. These studies analyze the influence of marital status on health and suggest that married men have less ill health, better psychological health, and adopt healthier lifestyles than non-married men, whereas the results for women are inconsistent (e.g.. Macintyre 1992; Wyke and Ford 1992; Benzeval 1998).

Further, these studies use mainly two different explanations. The 'health selection' thesis argues that unhealthy people are less likely to get married or more likely to divorce, whereas the 'social causation' thesis suggests a protecting effect of marriage due to material resources or social support (Benzeval 1998).

Following the 'social causation' thesis, Zick and Smith (1991) were able to find an association between marital status and mortality. They focus on women who became lone mothers due to divorce or widowhood and argue that people who are divorced or widowed have experienced extensive stress during the transition into lone motherhood, which may put them at a relatively greater risk of dying prematurely. However, other studies add restrictive perspectives on the advantages resulting from marriage. The results indicate that interpersonal stress in close social relationships may produce or exacerbate psychological distress in forms of depression, anxiety and loneliness (Mittelmark et al. 2004).

Recent studies concentrate less on marital status and compare partnered mothers (regardless of whether they are married or not) with lone mothers. The examined health outcomes differ in a great extent but can be subordinated in the two main groups of a) mental disorders and b) physical health.

The diverse mental disorders are mostly discussed with the different levels of stress between lone mothers and partnered mothers. Research examines higher prevalence of depressive symptoms (Cooper et al. 2008; Crosier et al. 2007; Lipman et al. 1997; Wang 2004); psychological distress (Franz et al. 2003); anxiety disorders (Afifi et al. 2006), and lower levels of well-being (Bull and Mittelmark 2009) for lone mothers compared to partnered mothers. In literature, stress is seen as the main determinant of lone mothers' mental disorders and is explained by the double burden faced by lone mothers due to sole financial responsibility and single-handed child care (Fritzell et al. 2007, Hope et al. 1999).

The results for physical health show that lone mothers have a higher risk for cardiovascular diseases (Young et al. 2005), chronic illness (Wickrama et al. 2006) and ill health (Roos et al. 2005) compared to mothers with a partner. Furthermore, studies report generally a worse health for lone mothers than for mothers with a partner (Curtis 2001, Westin and Westerling 2006).

These health disadvantages are mostly explained by the special economic circumstances of lone mothers. For example, Fritzell and Burström (2006) demonstrate in a trend analysis for Sweden that the prevalence rate of *less than a good* self-rated health increased for lone mothers during the 1990s compared to the 1980s. They explain the result as being caused by increased financial problems among Swedish lone mothers and outline that the economic strain has a substantial explanatory value.

This result corresponds with most other studies showing that both mental and physical health disparities between lone mothers and partnered mothers may be attributed to differences in certain socioeconomic factors (e.g. Curtis and Phipps 2004; Taragosz et al. 2003). Furthermore, Hope et al. (1999) showed a higher risk of psychological distress among lone mothers with financial hardship and unemployment. Additionally, there is also a higher prevalence of mental disorders for lone mothers who rely on welfare benefits (Coiro 2001). Further, the combination of poverty and the sole responsibility of childcare may culminate with social isolation, which is known as an important determinant of depression symptoms (Taragosz et al. 2003).

In addition, cross-national comparative studies analyze the health status in relation to welfare state arrangements. Fritzell et al. (2012) explore that lone mothers suffer financial hardship and worse health than partnered mothers irrespective of the different policy regimes seen in Italy, Great Britain and Sweden. On the contrary, Curtis and Phipp (2004) illustrate in a comparative analysis between Canada and Norway that Canadian lone mothers have a lower health status than partnered mothers, whereas there are no significant health differences were found in Norway. They attribute their findings to the more generous social benefits in Norway than in Canada which seem to absorb the socioeconomic differences between lone mothers and partnered mothers. Further, Burström et al. (2010) point out that the social context and social policies influencing the health of lone mothers could be the consequence of social position and environmental influences.

Unlike these studies, this chapter concentrates on lone mothers' health in a longitudinal perspective using panel data for the German population. We focus on determinants affecting mothers' health in the stressful period during the transition into lone motherhood and the first years of being a lone mother.

#### **Data & Methods**

#### *Data Management and Methods*

This study is based on data from the German Socio-Economic Panel (SOEP), a representative longitudinal dataset of the German population. The SOEP was initiated in 1984 and since then it has been conducted annually. It includes, among other things, detailed personal, social and economic information for all household members above the age of 16 (Wagner et al. 2007).

The SOEP contains all the necessary information to identify lone mothers (Hancioglu and Hartmann 2013). The data includes important information of household structure and socioeconomic factors as well as sociodemographic backgrounds of individuals. We define lone mothers as women who live with their underage child or children in a household without a partner. Our sample consists of respondents who were in lone motherhood at some point during the panel period between 1984 and 2011 (*N* = 2006).

By this definition, we observe more than one lone mother episode for some women during their life span. To identify the effects of transition into lone motherhood, we concentrate on the first episode. As a necessary condition of the data, we exclude left-censored lone mother episodes. That means that we only consider women who became lone mothers during the observation period, and do not consider women who were already lone mothers when they entered the study.

An additional requirement is that lone mothers with an observed transition into lone motherhood need to be observed for at least 3 years prior to the beginning of the lone mother episode. This restriction is of special importance because, according to the set-point model (Lucas 2007), measured outcomes can be affected by life-events (e.g. marriage, separation or death of partner) even certain years prior to the life-event. Therefore, we assume that health outcomes become worse just before the lone mother episode begins.

We take 3 years *(t−3)* before the lone mother episode begins into consideration (*t*0 is the first interview as a lone mother) to account for changes in health outcomes that might have occurred prior to the start of lone motherhood as well as to ensure a valid sample size. The lone mother episodes end with the end of lone motherhood or the end of the period of observation.

To explore individual health variation, we use longitudinal fixed-effects (FE) linear regression models. Health satisfaction and well-being are our dependent health variables which are measured on an interval scale and are linearly dependent on several predictor variables (Allison 2009, p.6). The notation of our basic model for *yit* is:

$$\mathbf{y}\_{\boldsymbol{\mu}} = \boldsymbol{\mu}\_{\boldsymbol{i}} + \boldsymbol{\beta}\_{\boldsymbol{\mu}} + \boldsymbol{\gamma} \boldsymbol{z}\_{\boldsymbol{i}} + \boldsymbol{\alpha}\_{\boldsymbol{i}} + \boldsymbol{\epsilon}\_{\boldsymbol{i}}$$

We have lone mothers (i = 1,…, n), which are measured at least for four times (t = 1, 2, 3, 4, …, n); *yit* describes our dependent outcome variables; *xit* represents variables which vary over time, and *zi* represents variables which do not vary over time. The use of fixed effects models reduces potential biases present in most crosssectional studies analyzing effects on health. These biases result from the inability to control for omitted characteristics which influence health status and are correlated with other characteristics (Ruhm 1996); *αi* and *ϵi* represent error terms, whereas *ϵi* is different for each individual at each time, *αi* varies across individuals, but not over time. For *ϵi*, there is a constant variance assumption made for all *i* and *t* (Allison 2009, p.6). With the assumption of statistical independence of *αi* and *ϵi* correlations between *αi* and *zi* are allowed2 .

#### *Variables*

The key measures that we use, and are available for all individuals in the SOEP data for all years, are self-assessed health satisfaction and self-assessed life satisfaction. The answers are recorded on a scale of 0 (poor) to 10 (excellent). The heath satisfaction variable closely corresponds to the measures of self-assed health (Frijters et al. 2005), which is commonly used as a predictor for assessing respondent's health status and is considered to be a good proxy for future morbidity and mortality

<sup>2</sup>For more detailed formal explanation of fixed effects methods see Allison (2009) and Wooldridge (2009).

(DeSalvo et al. 2006; Idler and Benyamini 1997). Life-satisfaction is a tool especially used in psychology. For example, Diener et al. (1999) and Lucas (2007) measure well-being by this cognitive measure, depicting satisfaction with life in general (Schimmack 2008).

The main independent variables of interest are two dichotomous variables. To distinguish between negative effects due to separation and effects which come along with being a lone mother, we generate a *separation dummy* and a *lone mother dummy.* Separation, which is attributed in the literature as a life-event (Lucas 2007) is derived from household composition. We identify a separation with a move-out from a joint household. For those lone mothers with a changed household composition, we create the *separation dummy* variable (1 = 1 year before separation). The aim of this variable is to check the robustness of the effects due to lone motherhood. We assume that the stressful situation begins shortly before the actual lone mother episode starts. The *separation dummy* allows us to capture those negative anticipation effects of separation.

To identify effects of lone motherhood and to consider the dynamics of lone mothers during the transition into lone motherhood, we generate a *lone mother dummy*. This variable measures years before the lone motherhood (= 0) and years in lone motherhood (= 1).

To use the potential of panel data, we use longitudinal information describing the duration of lone motherhood, assuming that changes in health outcomes can be explained by an increasing duration of lone motherhood. We also assume that stable (non-marital) partners may have an effect on lone mothers' health. Therefore we also consider lone mothers' partners who are not living in the same household.

Further, we control for sociodemographic and socioeconomic variables (see Table 15.1 for the overview of the control variables). We divide mother's age in six categories: (i) 17 to 25, (ii) 26 to 30, (iii) 31 to 35 (reference category), (iv) 36 to 40, (v) 41 to 45 and (vi) 46 to 61. Children's age is also distinguished into six categories following differences in institutional constraints. We consider the age of the youngest child: (i) new born: 0, (ii) pre-nursery from 1 to 3, (iii) kindergarten from 4 to 6, (iv) elementary school from 7 to 10 (reference category) and two categories for secondary schooling (v) from 11 to 15 and (vi) from 16 to 18 years. In addition, we control for the number of children in household because we assume that more than one child is associated with a higher burden. Both number and age of the children are included via a set of mutually exclusive variables indicating how many children are living in the household and how old the youngest child is.

We include two variables representing socioeconomic characteristics of the households – changes in net equivalent income and mothers' employment status. Changes in net equivalent income refer to the income of the respective previous year, and are categorized by its standard deviation3 . Five categories are used to describe changes: (i) major deterioration, (ii) small deterioration, (iii) constant level (reference category), (iv) small improvement and (v) major improvement. We distinguish employment status into five different categories: (i) full-time, (ii) part-time,

<sup>3</sup>The intervals are ascending: *x <* 1,5σ; −1,5 σ to −0,5*σ*; −0,5*σ* to 0,5*σ*; 0,5*σ* to 1,5*σ* and *x >* 1,5*σ*.

**Table 15.1** Variables


Source: SOEP 1984–2011 (pooled data)

(iii) marginal employment, (iv) vocational training and (v) not employed (reference category). Marginal employment is according to German social security law an employment relationship with a maximum monthly wage of 450€. The characteristic of being not employed includes besides individuals that are unemployed and looking for a job, also military and community service, maternal leave, and women in partial retirement who are not working anymore.

Further, we control for the utilization of (institutional) childcare, which plays a decisive role for lone mothers' labor market participation (Ott et al. 2011) and might have relieving effects on lone mothers' health. Childcare is controlled by a dichotomous variable which is only considered for women whose youngest child is younger than 6 years old.

#### **Results**

#### *Descriptive Analysis*

As a first step of analysis, we examine in Figs. 15.1 and 15.2 differences in *health satisfaction* and *well-being* between lone mothers and partnered mothers. Figure 15.1 shows the health satisfaction from *poor* (=0) to *excellent* (=10). The majority of lone and partnered mothers report their health satisfaction as good or excellent (from 7 to 10) whereas the percentages of moderate and poor self-reported health are relatively low for both groups. However, the figure shows that within the low values (from 0 to 5) depicting moderate or poor health satisfaction, lone mothers have a higher percentage than partnered mothers. In contrast, lone mothers have lower percentages than partnered mothers within the higher values (from 7 to 10), which illustrates in the sum a significant worse self-assessed health satisfaction of lone mothers compared to partnered mothers.

Figure 15.2 shows the well-being of lone and partnered mothers from *poor* (=0) to *excellent* (=10). Group differences are higher in Fig. 15.2 than in Fig. 15.1. Compared to partnered mothers, lone mothers are less likely to rate their well-being as good or excellent (from 7 to 10) whereas they rate their well-being more often as poor to moderate (from 0 to 6) than mothers with partners. As previously shown for health satisfaction, there is a very small percentage of mothers reporting poor wellbeing. However, in the sum lone mothers rate their well-being worse compared to partnered mothers. In general, the results are consistent with previous research. Lone mothers seem to be at a disadvantage regarding their health outcomes

**Fig. 15.1** Health satisfaction (SOEP 1984–2011 pooled data)

**Fig. 15.2** Well-being (SOEP 1984–2011 pooled data)

**Fig. 15.3** Health satisfaction and well-being over lone mother episode (SOEP 1984–2011)

compared to partnered mothers and group differences within well-being are higher than within health satisfaction.

In the second part of the analysis, we focus on lone mothers' health in a longitudinal perspective and consider changes prior to becoming a lone mother and during the lone mother episode. Figure 15.3 demonstrates the development of health satisfaction and well-being by comparing mean values over lone mother episodes and prior to lone mother episodes. The beginning of a lone mother episode is highlighted with a vertical line (t0).

Figure 15.3 shows that, even some years before women become lone mothers, there is a decreasing trend in both health outcomes. There is an especially strong decline in the mean values of well-being 1 year before separation. When observing mean values over time, two major development emerge. First, while there is a clear decreasing trend for health satisfaction, well-being values slowly increase and reach a plateau over the lone mother episode. It can be stated that Fig. 15.3 shows an increase of well-being values accompanied by separation (*t-1* and *t0*). And second, for health satisfaction the increase, which comes along with separation, is not as strong as for well-being; however there is a general decline of health satisfaction over the lone mother episode.

#### *Multivariate Analyses*

In the next part of analyses, we examine the effect of being in lone motherhood on health satisfaction and well-being after controlling for separation, duration of lone motherhood, squared duration of lone motherhood and mother's age. The aim of this longitudinal fixed-effects (FE) linear regression for the time period from 1984 to 2011 is to distinguish between the effects on health satisfaction and well-being due to separation (Amato 2010, Lucas 2007, Andreß et al. 2003), and the effects due to being a lone mother, which imply a special economic and social situation. To explore the robustness of the lone mother effect, we include all lone mothers with an observed separation.

Table 15.2 presents the fixed effects estimation for 688 women. The observation window varies for these women between 4 years (3 years prior to lone motherhood and 1 year in lone motherhood) and 27 years. The results coincide with the descriptive results, suggesting that separation affects mothers' health satisfaction and wellbeing (c.f. Figure 15.3) negatively. We find a strong significant negative effect of separation for health satisfaction and well-being.

Moreover, we also identify a strong significant negative effect of being in lone motherhood on well-being and a weak significant effect on health satisfaction. The results demonstrate that for well-being the effect of lone motherhood is even stronger than the ´separation effect`.

Further, the effects of duration in lone motherhood differ between the two health outcomes. Whereas the length of the episode has a significant negative effect on health satisfaction, the results show a positive effect of duration of lone motherhood on well-being. In other words, with an increasing length of the lone mother episode, the effects on health satisfaction are negative, but in contrast, the effects are positive for well-being.

In the second model (Table 15.3), we concentrate on changes within the transition into lone motherhood. Therefore, we consider 3 years prior to lone motherhood and the first 3 years in lone motherhood. In addition to model I (Table 15.2), we


**Table 15.2** Lone mother and separation effects (fixed effects regression)

Standard errors are reported in parentheses. \*\*\* p < 0.001 \*\* p < 0.01, \* p < 0.05 Source: SOEP 1984–2011

control for changes in net equivalent income and for the employment status describing the socioeconomic factors of the households. Further, we add the age of the youngest child, the number of children living in the household, the use of childcare, and the variable stable partner (who is not living in the same household) to illustrate a potential social support.

As shown in model I, we also identify negative effects on health satisfaction and well-being for the transition into lone motherhood due to separation and lone motherhood. Not surprisingly, there are positive effects of a stable partner on both health outcomes, since a new partner offers emotional support and can indicate the end of lone motherhood. Beyond this, a new partner could also be the reason for the separation from previous partner and could cause a soon end of the episode in lone motherhood. There are no effects for mothers' age, the age for the youngest child and the number of children living the household observable.

Table 15.3 examines how the socioeconomic characteristics that come along with a transition into lone motherhood affect health satisfaction and well-being. It shows that a loss in wealth is associated with a decline in lone mothers' well-being. This result is highly significant for major deterioration in the net equivalent income affecting lone mothers' well-being. Further, this model shows that a full-time and part-time employment have significantly positive effects on well-being. This might be a sign that women who actively participate in the labor market might be able to reach a certain level of financial independence.

Surprisingly, childcare has a negative effect on both outcomes. This effect can be explained with insufficient coverage of institutional childcare, especially in the 80s and 90s. Since, there is no information about the quantity and quality of childcare existing for all years; we are not able to differentiate between different possibilities of institutional and private childcare.


**Table 15.3** Health changes within the transition into lone motherhood (fixed effects regression)

Standard errors are reported in parentheses. \*\*\* p < 0.001 \*\* p < 0.01, \* p < 0.05 Source: SOEP 1984–2011

#### **Concluding Remarks**

Previous research has shown that lone mothers suffer disproportionally higher rates of physical and mental illness compared to partnered mothers. Our work reinforces previous findings that show group differences in health satisfaction and well-being between lone and partnered mothers. Unlike other studies, we focus on the identification of intra-individual health changes. In the first step of multivariate analyses, we differentiate between separation effects and effects due to lone motherhood. This differentiation is relevant because we argue that lone mothers' health is poorer than partnered mothers' health and this is not caused by a separation but is rather caused by the special economic and social situation of lone mothers.

Moreover, the estimation considers the duration of lone motherhood, which is associated with opposite effects on the health outcomes. We find a positive impact of a prolonged duration of lone motherhood on well-being. We interpret this result using the set-point model that suggests that after a life-event (here separation), wellbeing adjusts after a certain period of time and can reach the observed level prior to the separation. Lone mothers seem to adjust to their new social status and financial situation which affects positively their well-being as they deal with stress. In contrast, the opposite is true for health satisfaction as bringing up a child, alone and with reduced material and social resources, may affect health satisfaction. Therefore, we suggest that health worsens with the length of time someone has been a lone mother. This explanation follows the idea of ´social causation`. Even after using longitudinal data we cannot exclude the possibility that some lone mothers could have a lower chance of finding a new partner because of their poor health. This could be a reason for worsening health satisfaction over the duration of lone motherhood.

In the second model, we focus on the transition into lone motherhood and find significant effects of socioeconomic factors on health outcomes. The results highlight that a loss in wealth affects well-being negatively. The access to employment (full-time, part-time) can solve the problem of a major deterioration of income. However, we assume that lone mothers with full-time employment have other arrangements aside from institutional childcare. This assumption is confirmed with the result that childcare has a negative effect on health satisfaction and well-being. Access to universal childcare services for lone mothers in Germany should be a top priority for family policies. The implementation of a reliable and extended childcare (covering off-peak times) might bring an improvement in the health satisfaction and well-being of lone mothers and might also ensure the proper balance between work and family.

**Acknowledgements** This paper benefited from the support of the Swiss National Centre of Competence in Research LIVES – Overcoming Vulnerability: Life Course Perspectives, which is financed by the Swiss National Science Foundation (Grant number: 51NF40-160590).

#### **References**


Schimmack, U. (2008). *Measuring wellbeing in the SOEP*. In SOEP paper 145.


**Mine Kühn** is Research Scientist at Max Planck Institute for Demographic Research. Her research focuses on population health and family demography, with particular interest in developments of health inequalities across the life course. Dr. Kühn received her doctoral degree in Social Sciences from the Ruhr-University Bochum (Germany).

**Open Access** This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

## **Erratum to: Changing Lone Parents, Changing Life Courses**

**Laura Bernardi, Dimitri Mortelmans, and Ornella Larenza**

**Erratum to: Chapter 1 in: L. Bernardi, D. Mortelmans (eds.)**  *Lone Parenthood in the Life Course***, Life Course Research and Social Policies 8, https://doi.org/10.1007/978-3-319-63295-7\_1**

The original version of this book was published without Ornella Larenza listed among the chapter authors in Chapter 1.

It has been corrected and updated as:

Laura Bernardi, Dimitri Mortelmans, and Ornella Larenza

The updated online version of the chapter can be found at https://doi.org/10.1007/978-3-319-63295-7\_1