The aim of this paper is to analyze the relationships among income, financial strain, and child outcomes for different types of families in Germany. Almost 20% of German children grow up in a single-parent family or in a stepfamily (Federal Ministry for Family Affairs, Senior Citizens, Women and Youth 2013), and in particular, children in single-parent families and families with more than three children suffer from economic hardship or are at risk of becoming poor. In Germany, almost 20% of all children are at risk of poverty (Federal Ministry for Family Affairs, Senior Citizens, Women and Youth 2017), and such poverty during childhood affects economic wealth later in life (Bernardi et al. 2019; Bernardi and Mortelmans 2018). Therefore, to examine the inequality that might emerge from family structure, understanding how financial hardship affects child wellbeing is important. Additionally, knowledge about the relationships among economic hardship, financial strain, and child outcomes for children living in stepfamilies is limited. In stepfamilies, different types of children live in the household, including children common to the couple and stepchildren to one of the partners. Furthermore, if both partners bring children into the stepfamily, then each child has a stepparent. This diversity explains the complexity of stepfamilies.

The Family Stress Model (FSM) is well suited to explain the influence of economic hardship on the wellbeing of families and children. This model was developed by Conger et al. (1990, 1994, 2010) and predicts that “economic problems will lead to deterioration in marital relationships and increase the risk of marital instability” (Conger et al. 2010, p. 689). The FSM incorporates several aspects of economic strain. First, economic hardship is defined as having, e.g., a low income or other negative economic events. Economic hardship subsequently leads to the second aspect, namely, economic pressure, also called financial strain, which is viewed as, e.g., unmet material needs or financial cutbacks. In turn, a couple is affected by emotional and behavioral problems, which lead to the third aspect, namely, conflict, hostility, distance and withdrawal in the relationship. In the end, the quality and stability of the relationship is at risk (Conger et al. 2010). Ultimately, if a relationship suffers from financial strain and the relationship quality decreases, then child behavior may also be affected (Evans 2004; Neppl et al. 2015).

The strengths of the FSM include its ability to incorporate measurements of “behavioral interactions between spouses that might be influenced by economic deprivation” (Conger et al. 1990, p. 644) and to link economic strain, marital tensions and instability (Falconier and Epstein 2010). In a recent article, Masarik and Conger (2017) reviewed how the FSM has been applied in the past and how it has been extended, e.g., to introduce different structures, ethnicities and geographic settings. Interestingly, only a few studies, e.g., Schramm and Adler-Baeder (2012), have attempted to extend the FSM by acknowledging family structure and family complexity, thus not solely relying on two-parent households or single-parent households, as suggested by Barnett (2008). The difference between family structure and family complexity is that child status is more apparent when considering family complexity, which is important because children are most affected by separation. Furthermore, to clarify how each child is affected by economic hardship, understanding later life outcomes is also critical. Family complexity involves greater consideration of the types of children in a family, contributing to questions such as the following: In stepfamilies, do stepchildren differ in their outcomes compared to children who are common to both parents (Gennetian 2005; Halpern-Meekin and Tach 2008), and how are sibling relationships in such families (Baham et al. 2008)? Family structure takes the perspective of the household and distinguishes among different family types. Here, we provide a broader understanding of what type of family is most exposed to financial hardship. The most common distinction between family types can be identified by looking at nuclear families (two parents living with their biological child or children in one household), single-parent households (one parent living with his or her biological child or children in one household) and stepfamilies. Different definitions of stepfamilies exist (Juby et al. 2001; Ganong and Coleman 2004; Pryor 2008; Heintz-Martin 2013), the most prevalent of which asserts that in a stepfamily, one of the parents is not the biological parent of the child or children. A stepfamily is called a stepmother or stepfather family according to the biological parent of the child or children, whereas a stepfamily is called a stepfather/stepmother family if both parents bring a child into the union. Furthermore, the so-called blended stepfamily is a special type of stepfamily because on one hand, at least one child is not the biological child of one of the partners, while on the other hand, the new couple has at least one common child together. We follow Juby et al. (2001) and call these families blended stepfamilies.

Research examining income, subjective financial strain and child outcomes simultaneously has been limited because data measuring all three aspects are scarce, especially in Germany. We try to fill this gap by applying the FSM in two ways. First, we extend the model by not solely focusing on family structure but rather focusing more on family complexity. Second, we include child behavior as an outcome of financial strain, an extension that has already been proposed by Conger and Conger (2002) and Masarik and Conger (2017). This paper aims to first analyze which type of family is mostly affected by economic hardship. Second, we test which type of family experiences more financial strain. Third, we examine the consequences of financial strain for children with respect to behavioral problems.

The German Context

Similar to most Western families, German families are increasingly diverse with regard to their household composition; consequently, family types are distributed differently than in the past. No official statistics are available on the distribution of stepfamily types in Germany, but data from several surveys are available to estimate the percentage of stepfamilies. For example, estimations from the survey “Growing up in Germany”: AID:A II wave reveal that approximately 79% of families are so-called nuclear families (two biological parents and their children), 10% are stepfamilies (one parent is the biological parent to the child(ren) and the other is a stepparent), and 11% are single-parent families (Federal Ministry for Family Affairs, Senior Citizens, Women and Youth 2013). Germany is still characterized by a more traditional model of the division of labor in couples, and full-time working women with young children are still rare. Most women with small children work part time (Andreß et al. 2006; Zabel and Heintz-Martin 2013). The male breadwinner regime remains prevalent according to Lewis et al. (2008); thus, separation is a risk factor for becoming poor. Consequently, women are at a high risk of experiencing economic hardship if their marriages fail.

Some financial help is available for low-income families in Germany, which also applies to single mothers. If their income is too low or if they are unemployed, they receive benefits to cover their subsistence costs (Bröckel and Andreß 2015). Each family, regardless of income, has a child allowance of approximately 190€ per child per month.

Some of the tax benefits that parents with children receive are not applicable to single parents because their income is too low to benefit from such benefits, which is especially true for women living only on welfare tax benefits because they do not receive a taxable income (Neuberger et al. 2019). Single parents also receive payments from the youth welfare authorities for their children if the absent parent it not willing or able to pay his or her part of the child-rearing costs that he or she is normally obliged to pay (Bröckel and Andreß 2015). In addition, efforts have been made to push (single) mothers back into the labor force through expansion of day care availability, as well as the law change in 2008 requiring men to pay child support only until the youngest child is 3 and not 16 years old as before (Bröckel and Andreß 2015). Despite these efforts, single mothers are still exposed to poverty in Germany (Neuberger et al. 2019).

Additionally, the joint tax system benefits women with lower earnings rather than their husbands staying at home. Furthermore, school hours and work hours lack coordination, and full-time day schools are not as frequently available as they should be according to Büchel and Spiess (2002). Andreß et al. (2006) showed in a comparative study that women experience significant income loss at the time of separation; however, this is not the case for men. Interestingly, the findings of the authors revealed that German women recover from this income loss within twelve years of their separation only if they have a new partner; otherwise, they need 18 years to recover. This result may indicate that forming a stepfamily not only helps women leave financially difficult situations but also benefits their children. Nevertheless, single parents are well known to suffer the most from economic hardship compared to other family types (see also Kreyenfeld and Martin 2011).

Even three decades after the reunification of East and West Germany, differences between the two parts of the country remain. These discrepancies are not solely economic, such as the higher level of unemployment in eastern Germany and economically deprived populations in some regions. Differences between eastern and western Germany also remain with respect to the distribution of family types (see also Kreyenfeld and Martin 2011). The higher prevalence of stepfamilies and single-parent families in eastern Germany can be explained by the higher rates of nonmarital birth in eastern Germany because the parents involved have a higher risk of separation. For example, in 2012, almost 60% of all births in eastern Germany were out of wedlock compared to 28% in western Germany (Kreyenfeld et al. 2017).

Conceptual Framework and Hypotheses

Most previous research on economic wellbeing agrees that socioeconomic status (SES) variables, such as income, education and work status, are good measurements for economic wellbeing (Bradley and Corwyn 2002; Ensminger and Fothergill 2003; Conger et al. 2010). Research on financial strain found strong connections between the financial burdens that people experience and variables such as income, financial literacy, living arrangements, and education (Malone et al. 2010). Additionally, the effects of socioeconomic hardship on children have been previously documented (Kiernan and Huerta 2008). Conger et al. (2010) also discussed the relationship between financial stress and decreasing marital quality. Poorer martial quality can increase the risk of separation. In contrast, wealthier couples have increased levels of happiness and satisfaction with their partnerships (Karney and Bradbury 2005).

The FSM has often been applied to explain the influence of economic hardship on the wellbeing of families and children (see, e.g., Kinnunen and Feldt 2004; Aytaç and Rankin 2009; Neppl et al. 2015). We build our hypothesis upon the FSM and extend the model by examining not only family structure but also family complexity to better understand what type of child is more affected by financial strain. We start by examining the roles of economic hardship and financial strain in different family types and then explore the effects of economic strain on children by considering family complexity.

Economic Hardship in Single-Parent Families and Stepfamilies

Economic hardship refers to people with a low income, which is defined as being below the poverty threshold. Children in single-parent families are assumed to accumulate less human capital because one parent has fewer resources to support such accumulation than do two parents (Lopoo and DeLeire 2014). Additionally, women in single-parent families are at a high risk of financial hardship, which has been shown to also affect the wellbeing of these women (Stack and Meredith 2018).

The separation (or death) of a parent can lead to stress for a child, reduce his or her economic wellbeing, affect his or her emotional and physical development and lead to lower school performance (see, e.g., Lopoo and DeLeire 2014). The results of the study of Lopoo and DeLeire (2014) are noteworthy because they show two major aspects of the outcomes of divorce for children. First, their results suggest that children in remarried families are better off than children in single-parent families; more notably, however, in the long run, such children are even better off than children living continuously with two married parents. The authors demonstrated that economic wellbeing is reduced for children after a maternal separation; however, when they accounted for family size, their results suggest that remarriage is a protective factor. “All children have greater resources available to them later in childhood, but children of mothers who have a dissolution and remarry have more resources available per capita than children of continuously married or the children of never remarrying mothers” (Lopoo and DeLeire 2014, p. 41). The authors argue that mothers who remarry may be different from those who do not remarry; however, they were not able to control for this factor. The authors suggest further investigation into the presence of any selectivity between those two groups, e.g., mothers with fewer children are more likely to remarry than mothers with more children. However, in contrast to the results of Lopoo and DeLeire (2014), other studies suggest that stepfamilies have fewer economic resources available compared to nuclear families (e.g., Thomson et al. 1994; Stewart 2001). For example, Ribar (2004) suggests that stepparents may invest less in a stepchild because they do not value him or her similarly to a biological child. Studies considering stepfamily-specific stressors related to the relationship between economic pressure and marital quality are rare.

One of the first studies focusing on the effects of economic pressure on men and women in stepfamilies was carried out by Schramm and Adler-Baeder (2012). Perhaps women who remarry are more likely to reenter the workforce because with a partner in the household, reconciliation between work and family may be easier. In addition, women who marry for the second time may look for a man who can support a family financially rather than a partner who is not able to contribute to the family income, assuming that he is willing to do so.

In addition to the fact that children living in remarried families may be protected from economic hardship, research has shown that these children are more at risk for negative school outcomes (see, e.g., Astone and McLanahan 1991) and for having emotional problems compared to children growing up in nuclear families (Astone and McLanahan 1991; Coleman et al. 2000; Ribar 2004). Lopoo and DeLeire (2014) suggest that these negative outcomes may not be attributed to a reduction in parental resources but rather the stress that dissolution and remarriage causes children.

Our first hypothesis emphasizes the association between income and different family types, namely, nuclear families, single-parent families, stepfamilies and blended stepfamilies. Given the high prevalence of single parents being at risk of being poor and following the discussion in the previous section, we anticipate the following:

H1

Single-parent families are more at risk of being below the poverty threshold than are nuclear families, stepfamilies or blended stepfamilies.

As argued above, blended stepfamilies have been found to be similar to nuclear families regarding certain characteristics, such as socioeconomic factors; therefore, we expect blended stepfamilies to be similar to nuclear families.

The Importance of Feeling Financial Strain

Measuring economic hardship not only via objective measures, such as income or educational attainment, but also by incorporating subjective impressions (e.g., White and Rogers 2000) seems to be promising as this approach considers both aspects of financial wellbeing, namely, the subjective and objective aspects. Financial satisfaction or, for the negative connotation, economic pressure can be viewed as one’s own evaluation of his or her financial situation. Predictors of financial satisfaction have been argued to be important in the context of family economics (see Joo and Grable 2004). This framework should incorporate direct and indirect effects, as well as objective and subjective measurements. What does financial wellbeing mean? Financial wellbeing is defined as “a state of being financially healthy, happy, and free from worry” (Zimmerman 1995) and is related to subjective appraisal of the personal financial situation (Joo 2008, p. 22). Joo and Grable (2004) argue that financial satisfaction refers to “contentment with one’s material (objective) and nonmaterial (subjective) financial situation” (p. 27). Researchers have tried to measure financial satisfaction in different ways. Some argue that it can be measured by a single question, i.e., “how satisfied are you with your financial situation?” (Morgan 1992, p. 127), or as Greenly and colleagues ask, “how comfortable and well-off are you financially?” (Greenley et al. 1997, p. 251). Other researchers favor multiple-item measures incorporating, e.g., financial adequacy, perceived economic wellbeing, and satisfaction with one’s level of living (e.g., Draughn et al. 1994). Falconier and Epstein argue that “economic strain is an individual’s subjective evaluation of his or her financial circumstances” (Falconier and Epstein 2010, p. 782), which refers to whether people view themselves as financially burdened.

Joo and Grable (2004) suggest six spheres to incorporate into the measurement of financial satisfaction. First, demographic and socioeconomic characteristics, such as gender, education or income, should be considered. Second, financially stressful events in people’s lives, including illness, investment losses or life cycle events, such as divorce or the death of a partner, should be evaluated. Third, financial behavior, such as paying bills or comparison shopping, should be assessed. Fourth, financial solvency, which refers to having better ratios, should be considered. Fifth, financial attitudes, such as the personal perception of one’s own cash management, and last but not least, financial knowledge, which is also called financial literacy and refers to the basic knowledge of one’s financial belongings, should be incorporated.

Malone et al. (2010) demonstrated that in the United States, single women, women cohabitating with a partner, and women in stepfamilies have more concerns about their financial situation than married women in nuclear families. For women in stepfamilies and single-parent families, knowledge of their financial situation is even more crucial because they face various financial issues, such as child support, alimony, their own retirement planning, and a potential new partner’s financial planning and financial possibilities. Married and remarried couples may be very different in their financial planning. First, if at least one partner witnessed a separation and therefore a period of single parenthood, then he or she consequently had to reorganize his or her financial belongings. When entering a stepfamily, partners must discuss whether to pool their money, whether the new partner is willing to share his or her money to care for his or her new stepchildren, and how they will organize their financial future in terms of retirement planning, savings or investments, and day-to-day money. All of these decisions may also depend on the possibility that the new partner must pay child support to his or her ex-partner for children who are no longer living with them (Malone et al. 2010).

Interestingly, although stepfamilies have complicated financial situations that may be problematic to manage, most do not discuss such arrangements prior to remarriage, and women may not be as involved in these financial issues as they should be according to Coleman and Ganong (1989). Some researchers argue that women should strive to be financially independent, that women tend to view their own financial situations more negatively than positively, and that they are less risk-tolerant than men (Anthes and Most 2000; Grable 2000; Loibl and Hira 2007; Zagorsky 2003).

We expect higher rates of feeling subjective strain among single-parent families and stepfamilies than those among nuclear families and blended stepfamilies because one parent is often solely responsible for financial issues in the former family types. We also expect that single-parent families and stepfamilies feel a strong subjective financial strain regardless of their objective incomes. Thus:

H2

Single-parent families and stepfamilies are more at risk of feeling subjective financial strain than are nuclear families and blended stepfamilies.

Following the FSM, we also consider satisfaction with a relationship, i.e., the amount of conflict in the relationship and marital status. Because conflict in a partnership can affect the psychological wellbeing of people, we included measurements of mental health. Accordingly, we excluded single-parent households and controlled for marital status, satisfaction in the relationship and conflict in the relationship; thus, we anticipate the following:

H3

Stepfamilies are more at risk of feeling subjective financial strain than are nuclear families and blended stepfamilies.

Effects of Economic Strain on Children and the Role of Family Complexity

The FSM anticipates that a negative parental relationship due to constant economic hardship extends to the parent–child relationship. Parents under stress may be harsher, uninvolved, or inconsistent in child rearing (Conger and Conger 2002).

We therefore want to discuss the relationship between economic hardship and child outcomes operationalized via the Strength and Difficulties Questionnaire (SDQ) (Goodman 1997). Second, we want to obtain a deeper understanding of family complexity by analyzing what type of child is more affected by financial hardship and thus exhibits negative behavior. Beyond family type or in other words, family structure, the importance of family complexity has emerged. Research on family complexity with regard to financial strain remains limited (Schramm and Adler-Baeder 2012). Therefore, we try to fill this gap by considering not only family structure and thus family type but also family complexity by acknowledging the status of a child. This approach seems productive to us because families that differ from the nuclear family may encounter more financial issues and experience more burdens because of their compositions, as we show in Fig. 1. Additionally, this approach enables us to determine whether differences exist in child behavior between stepchildren or common children in a stepfamily.

Fig. 1
figure 1

(Adapted from the Family Stress Model by Conger et al. 1990)

Conceptual family type and family stress model.

Full size image

Brown et al. (2015) argue that the study of child complexity is still in the beginning stages because of the data available, and that previous studies have often focused on adolescents rather than children. Second, the issue of economic wellbeing has largely been ignored. Their results highlight the necessity of focusing on family complexity and show that economic wellbeing is linked to family complexity, i.e., the presence of half-siblings has a negative effect on economic wellbeing. A stepparent may be likely to invest more resources into his or her own biological child than a stepchild (Biblarz and Raftery 1999; Conger et al. 1990; Evenhouse and Reilly 2004). However, with regard to SES, researchers have suggested that a stepparent provides financial resource support and helps monitor the child (e.g., Erola and Jalovaara 2017; Sweeney 2010). All of these processes of adaptation are strongly dependent on the ages of the children and how long the children lived with only one parent before the stepparent joined the family (Heintz-Martin 2013).

As previously discussed, children raised in nuclear families are known to have more economic, social, and cultural resources at their disposal compared to children living in single-parent families or stepfamilies (Erola and Jalovaara 2017), although contradictory findings have been reported, such as those of Lopoo and DeLeire (2014). Children not living with two biological parents have been documented to fare worse in terms of psychological wellbeing and tend to have more behavioral problems and health issues, as well as poorer education and labor market attainment (Härkönen et al. 2017). However, in addition to such economic, social, and cultural impacts, how does financial strain experienced by mothers affect child behavior, and how are children differently affected by such strain? In other words, what is the significance of family structure? The issue of family structure or family composition has not been largely examined (see for discussion, e.g., Härkönen et al. 2017) mostly due to the lack of available data for such complexity. Our data allow us to disentangle each child’s status. We are therefore able to examine whether, e.g., stepchildren are more likely to have behavioral problems compared to children living in single-parent families.

Children experiencing parental separation and entering a stepfamily face more transitions than children living in nuclear families or children living permanently in single-parent families. Findings regarding the impact of transitions on child outcomes suggest that the type of transition is relevant to both children and their outcomes later in their lives (Härkönen et al. 2017).

Previous research has consistently shown that children in stepfamilies have wellbeing outcomes similar to those of children in single-parent families (see for discussion, e.g., Amato 2001). Biblarz and Rafferty summarized several theoretical assumptions regarding why children in single-parent families or stepfamilies have lower attainments than children from two continuously married parents (Biblarz and Raftery 1999). Nevertheless, others report that family structure has only a modest effect on child achievement (Gennetian 2005). Why do stepchildren and children of single parents often have poorer outcomes?

Children in stepfamilies face special challenges compared to children in nuclear families and similar but additional challenges to those of children in single-parent families. First, similar to children in single-parent families, children in stepfamilies must maintain a relationship with their other biological parent with whom they are no longer living or who they see on a regular or irregular basis depending on the arrangement agreed upon by parents. Therefore, they have “day-to-day” transitions to address, and related conflicts may occur with such arrangements between the parents. Second, a circumstance unique to stepchildren is dealing with a new parent, which may not always be easy and often depends on the age of the child at the time of the formation of the stepfamily (Heintz-Martin 2013). Additionally, a stepchild may become a stepsibling if the new parent brings children into the family, and he or she must adapt accordingly. Third, if a new child is born to the stepfamily, then the child becomes a half-sibling, which adds further complexity to the family and requires further adaptation.

Brown and colleagues (Brown et al. 2015) emphasize in their study the relationship between family complexity and child wellbeing and address the issue of sibling composition in complex families. The authors argue that family structure does not sufficiently emphasize family complexity with regard to child outcomes. By family complexity, they refer to sibling composition, which applies to all family types but is especially noteworthy for stepfamilies and blended stepfamilies. In addition, as stated before, children in nonstandard families must adjust not only to new parents but also to stepsiblings as well as half-siblings if a common child is born to the couple (in other words, transformation into a blended family). Previous studies have suggested that family complexity is negatively associated with child wellbeing (Gennetian 2005; Halpern-Meekin and Tach 2008) and that children in complex families have a lower “income-to-needs ratio” and lower economic wellbeing (Brown et al. 2015). Previous research suggests that children from nuclear families have better educational outcomes than children from other family types (Ginther and Pollak 2004). Interestingly, Ginther and Pollak (2004) showed that some of the effects of the poorer outcomes of single-parent children disappear if controls such as the income or educational status of the mother are included in the analyses.

Considering the above discussion on stepfamilies and single-parent families and their special situation compared to that of continuously married or cohabitating parents, we argue that the wellbeing of children in stepfamilies and single-parent families is affected if parents experience financial strain. The outcomes of the children would be visible by their exhibiting negative behavior. To acknowledge not only family structure but also family complexity by considering child type, we anticipate the following:

H4

Stepchildren in blended stepfamilies and children in single-parent families are more likely to have behavioral problems than are children in other family types.

Methodology

Data and Sample

Our analyses are based on data from the second wave of a large, representative German survey on children and youth, namely, those aged between 0 and 32 (“Growing up in Germany”: AID:A II wave, 2013–2015) (Walper et al. 2015). The sample was randomly drawn from a nationwide German population register, and participants were contacted by professional interviewers. Standardized interviews with the target person (from age 9 upwards), with a primary caretaker for minors—usually the mother and (if applicable) her partner—were conducted by telephone. The AID:A II survey gathered information about the lives of children, teenagers, and adults, including their stresses and strains, their family structures, their socioeconomic situations, and their subjective financial strains. The initial sample included 22,424 respondents. Questions related to socioeconomic issues, such as living conditions, education, and employment status, were all answered by the mother and by the father in some cases. We restricted our sample solely to mothers due to the small number of male respondents.

The analytical sample included 12,561 target children aged between 0 and seventeen and living in the household, which allowed us to identify the following family formats: nuclear families, single-parent families, and stepfamilies (defined as two-parent families in which at least one child in the household is not the biological or adopted child of one of the parents). As mentioned above, stepfamilies can be divided into stepfather families, stepmother families and stepmother/stepfather families. Due to the low number of cases in the categories of stepmother families and stepmother/stepfather families, we conducted the analysis using only the broader category of stepfamilies. We further identified so-called blended stepfamilies (Juby et al. 2001), which refer to either a stepfather and/or stepmother family with a common child.

Method

The models that we used were ordered probit models because our dependent variables are in categorical order and the distributions are standardized to normal. The determinants of low income for family types were investigated in model 1, and in models 2 and 3, the determinants of subjective economic strain for family types were investigated. Finally, in our last models (model 4 and model 5), the SDQ was the dependent variable.

The dependent variable for H1 was a status below the poverty threshold. The variable was coded as 0 for an income above 60% of the poverty median threshold of approximately 20,000€/year and as 1 for an income below 60% of the median.Footnote 1 The indicator poverty risk is based on households’ net per capita income weighted by household needs (indexed by household composition according to the new OECD scale). The threshold value for poverty risk is set by the EU at 60% of the median needs-adjusted equivalence income, thus dividing the sample into two groups (0 = above the poverty threshold, 1 = below the poverty threshold).

The dependent variable for H2 and H3 was subjective financial strain. In our study, the subjective indicator for economic pressure was measured by a respondent’s estimation of her own experience of economic hardship.

The respondents were asked to rate the following three statements related to their financial situations: “We have enough money for everything that we need”, “We often have to pass on something because we must limit our finances”, and “In our family, money is often short”. These variables were combined to establish an indicator for financial strain, which was coded into three groups: 1 = “feeling a small financial strain (below 25%)”, 2 = “feeling somewhat of a financial strain (medium 50%)”, and 3 = “feeling a strong financial strain (above 25%)”. Subjective economic deprivation has been viewed as a robust parameter (Conger and Conger 2002).

For the last hypothesis, H4, the problem behavior of the children was the dependent variable. We used the SDQ, which is a method used to study child and youth problem behavior (see Goodman 1997) and the depressive symptoms of respondents. Following the cut-off points for the SDQ (see www.sdqinfo.org), we used two categories: 1 = behavioral problems and 2 = no behavioral problems. Notably, in our data, children aged nine to 17 reported their own problem behavior, while mothers reported the problem behavior of children aged four to eight. For this set of analyses, we only included the children who reported their own behavioral problems; therefore, our sample size for this set of analyses is smaller than the samples for the previous analyses.

Main Explanatory Variables

The family type variable consists of the categories of nuclear family, single-parent family, and stepfamily. Stepfamilies include all of the types that we have previously discussed, namely, stepfather families, stepmother families, stepmother/stepfather families, and so-called blended stepfamilies, which refer to families with both a common biological child and stepchildren (Juby et al. 2001; Heintz-Martin 2013).

In our sample, we identified children living in a specific family type and not in a household structure. In western Germany, 86% of children lived in nuclear families, 8% lived in single-parent families, and 6% lived in stepfamilies. For eastern Germany, we found that 83% of children lived in nuclear families, 9% of children lived in single-parent families, and 8% of children lived in stepfamilies (χ2 = 24.69; p < 0.001).

Covariates

Region

One main explanatory variable was region; here, we differentiated between western Germany and eastern Germany. Even three decades after the reunification, differences between eastern and western Germany remain prevalent, mostly with regard to income, the work status of women, and the number of stepfamilies (Kreyenfeld and Martin 2011).

Educational Attainment

We also controlled for educational attainment following the Comparative Analysis of Social Mobility in Industrial Nations (Casmin) (see König et al. 1987). The variable was coded = 1 for some secondary education with or without vocational training, = 2  for a college degree with or without vocational training, and = 3 for a university degree.

Employment Status

Employment status was added to the model with the categories of working full time, working part time, and unemployed.

Age

The age of the respondents and the youngest child were included as continuous variables. The mean age of the mothers was 40.74 years, and the mean age of the youngest child was 6.93 years.

Number of Children

The number of children living in the household was added to the model. The variable was coded = 1 if one child was living in the household, = 2 if two children were living in the household, and = 3 if three or more children were living in the household.

Conflict in the Partnership

In Model 3 in which we tested only families headed by two parents, we also added the level of conflict in the partnership (we summarized four categories into two categories, conflict versus no conflict) and satisfaction regarding the relationship (ranging from 1 = very happy to 6 = not happy at all, then summarized into three categories ranging from 1 = satisfied to 2 = okay and 3 = not satisfied).

Marital Status

We distinguished between married and cohabitation because some two-parent families are living in cohabiting unions.

Results

In the following section, we present some descriptive findings, followed by a discussion of the ordered probit models.

Descriptive Findings

The largest proportion of families includes target children living in nuclear families (86%), followed by target children living in single-parent families (8%) and target children living in stepfamilies (6%). The relative percentage of stepfamilies may be small because the sample includes only target children below the age of 17, see Table 1.

Table 1 Descriptive statistics for different family types in % (N = 12,561)
Full size table

In eastern Germany, the percentage of target children living in a single-parent family or stepfamily was higher compared to that of target children living in the same family formats in western Germany. These results are consistent with those of other German studies on stepfamilies (see Kreyenfeld and Martin 2011).

Interestingly, women in nuclear families held the highest educational degrees (41%) compared to women in single-parent families (31%) and stepfamilies (24%).

Women in single-parent families and stepfamilies had a high prevalence of secondary education (35% and 42%) compared to women in nuclear families (30%). This outcome may play a role in such women’s higher levels of feeling financial strain and their lower incomes. With regard to the working situation, the well-known relationship between German women and work status was reflected in the results (Hipp et al. 2015; Zabel and Heintz-Martin 2013), i.e., most women worked part time (50% in nuclear families, 47% in single-parent families, and 43% in stepfamilies).

Regarding income, compared to nuclear families, higher percentages of single-parent families and stepfamilies (31% and 15%, respectively) had a low income.

The mean age of the women in the sample was approximately 41 years. The mean age of the children was approximately 7 years. Unsurprisingly, stepfamilies were more likely to have three or more children (43%) than were nuclear (24%) or single-parent families (13%). Regarding the distribution of feeling financial strain, we can see that nuclear families seem to have had the lowest overall percentage of feeling some strain (51%) or feeling a strong strain (9%) compared to single-parent families feeling a strong strain (24%) and some strain (54%) and stepfamilies feeling a strong strain (15%) and feeling some strain (52%). The respondents were also asked about their mental health. Overall, respondents in stepfamilies reported a slightly higher likelihood of being in a bad mood, which may be explained by the fact that women in such families still bore the main burden with regard to household tasks (Heintz-Martin et al. 2015).

As explained above, we also tested the SDQ as a dependent variable. In the descriptive results, one can see that target children in single-parent families had behavioral problems somewhat more often (10%) than children in both nuclear families (6%) and children in stepfamilies (9%).

We calculated bivariate correlations for all of our models (Table 7 in the “Appendix” section), and the results show similar patterns for the poverty threshold and financial strain: Mothers in eastern Germany were more often below the poverty threshold and felt more financial strain; the same was true for older, less-educated women, and women who did not work. As expected, mothers with more than one child had a high risk of living below the poverty threshold and reported more feelings of financial strain. Only the age of the youngest child was solely connected to the poverty threshold. No significant correlation with financial strain was found. Although correlations existed between the more psychological factors (conflict in the partnership, satisfaction with the relationship, depressive symptoms) and the poverty threshold, these correlations were all lower than those between the psychological factors and financial strain. Significant relations existed between children’s reported behavior problems and both the poverty threshold and financial strain; however, the magnitude of these correlations was not very high.

Multivariate Results

Regarding the ordered probit models used to test our first hypothesis, we expected that single-parent families are more likely to be below the poverty threshold than are nuclear families, stepfamilies or blended stepfamilies. Table 2 shows the results and supported our first hypothesis; we can see that single-parent families and stepfamilies fared much worse than nuclear families. Eastern Germans were also financially disadvantaged compared to western Germans. The results suggest that higher education lowers economic hardship. Unemployment also had a significant negative effect on the economic situation.

Table 2 Ordered probit model predicting if household is below the poverty threshold (Model 1)
Full size table

Regarding our second hypotheses, Table 3 shows that single-parent families and stepfamilies were more at risk of feeling subjective financial strain than nuclear families and blended stepfamilies.

Table 3 Ordered probit model predicting if respondent experiences subjective financial strain (Model 2)
Full size table

We can see that single-parent families were more at risk of reporting subjective financial strain, followed by stepfamilies. Respondents in eastern Germany reported more financial strain than their counterparts in western Germany. Education was positively associated with reporting lower feelings of financial strain. Regarding age, the results suggest that older women suffered less from financial strain than younger women. For the ages of the children, the results were not significant. However, when we added the age of a child as a categorical variable, the results for the youngest group (aged 1–3 years) were significant (the results are not shown). This outcome may be explained by the fact that women with young children (aged 1–3) were often not working and were no longer eligible for the “Elterngeld” (nor were mothers of children aged 0); hence, these women were more likely to report financial stain. The number of children in the household significantly increased the financial strain. In addition, having an income below the median supported the feeling of a strong subjective financial strain. Work status was not significant.

Overall, the findings supported our second hypothesis. Single-parent families and stepfamilies were more at risk of experiencing financial strain than nuclear families, which was also true when controlling for the objective financial situation. One can say that single-parent families, followed by stepfamilies, were more sensitive to feeling financial strains.

Table 4 shows the results for model three in which we examined only families with two parents in the household to determine whether financial strains affected couples’ partnership satisfaction. As shown above, our results again suggested that stepfamilies were the most affected by financial strain. In addition, our results suggested that the level of conflict in partnerships and the level of satisfaction in the relationship were important for the feeling of financial strain, i.e., more conflicts and less satisfaction corresponded to a higher likelihood of respondents reporting financial strain.

Table 4 Ordered probit model predicting if respondent experiences subjective financial strain in two parent families (Model 3)
Full size table

Furthermore, the results supported the assumption that blended families are more likely to be similar to a nuclear family in terms of outcomes, i.e., a stepfamily that is getting along well may be more likely to decide to have a common child and therefore be more similar to a nuclear family in its outcomes (see for discussion Heintz-Martin 2013).

Stepfamilies were more exposed to financial strain compared to nuclear families or blended families, supporting our third hypothesis.

In our third set of analyses in which we examined how the financial situation (objective and subjective) correlated with the psychological wellbeing of children and attempted to obtain a deeper understanding of what type of child in a household was more affected by financial strain, our dependent variable was the SDQ. In this model, we acknowledged family complexity by examining child types. Therefore, we disentangled our family structure variable in detail to detect each child and his or her status in the family.

The fourth hypothesis stated that stepchildren in blended stepfamilies and children in single-parent families are more likely to have behavioral problems than are children in other family types.

As shown in Model 4 (see Table 5), single-parent families and stepfamilies had a higher risk of having children with an SDQ indicating problem behavior.

As expected, the risk of having children with behavioral problems was higher if parents reported feeling a strong financial strain.

Thus, we found support for our fourth hypothesis, i.e., children in single-parent families and stepchildren in blended stepfamilies have the highest risk of exhibiting problem behavior. Additionally, the coefficient for stepchildren was not significant but was in the same direction.

The other variables that we controlled for will not be explained further because they exhibited the same direction as those in the previously discussed analyses

Table 5 Ordered probit model predicting if child reports problem behavior (Model 4)
Full size table

In a last Model (Model 5), we explored the interaction between child status and experiencing financial strain (see Table 6). The rationale behind this model was to test whether the effects of financial strain were different for children in different families. The results clearly suggested that the strongest effects of financial strain on problem behavior applied to stepchildren in blended stepfamilies

Table 6 Ordered probit model predicting if child reports problem behavior with interaction between child type and financial strain (Model 5)
Full size table

These results highlighted the importance of distinguishing not only between family types but also between the types of children and showed that significant differences existed in SDQ outcomes depending on how financial strain was experienced.

Summary and Conclusion

This paper focuses on two aspects regarding family type and economic wellbeing. On the one hand, we examined how financial strain differs among family types; on the other hand, we examined how financial strain affects child wellbeing in different family types. We analyzed single-parent families, nuclear families and stepfamilies, as well as blended stepfamilies, by applying ordered probit models using a German data set, which allowed us to consider not only family structure but also family complexity. This approach was especially helpful for understanding what type of children may suffer more and how financial strain affects child behavior to disentangle the type of child involved, e.g., a stepchild or a common child in a blended stepfamily.

Our results are consistent with those of previous studies and show that single-parent families (Amato 2010; Stack and Meredith 2018) followed by stepfamilies are the most exposed to poverty (see, for example, Biblarz and Raftery 1999; Kreyenfeld and Martin 2011). Our findings highlight that this outcome is true not only for objective poverty measured via the poverty threshold but also for the subjective feeling of financial strain. Interestingly, the differences in experiencing financial strain between family types remain when controlling for the objective financial situation measured by income, indicating that mothers in single-parent families and mothers in stepfamilies, being in the same financial situation, feel more strain than mothers in nuclear families in the same financial situation. One explanation for this outcome for women in stepfamilies and in part for single mothers may be that they must maintain a relationship with their new partners and their children and may still have issues, in some cases, with the biological fathers of their children. Considering all factors together, this can lead to a feeling of exhaustion for women. Overall, women are still the “pivotal figures that hold the glue together” (Ganong and Coleman 2004, p. 109), which may be exhausting and lead to higher scores in depressive symptoms, although they have not been found to be significant (Table 6).

If women suffer financial strain, regardless of their family types, then they may be under more stress, which can influence their children; therefore, these children are more at risk of showing problem behavior. However, we can see that differences exist between family types in children in single-parent families and stepchildren in blended stepfamilies, with stepchildren being more likely to show problem behavior than children living with two biological parents. One could argue that stepchildren and children in single-parent families seem to be more sensitive to the stress that they witness, which is defined here as financial strain, compared to children from other family types. In concordance with the FSM elaborated by Conger et al. (1994), our results show that in two-parent families, relationship quality is important for the feeling of financial strain, i.e., a good-quality partnership seems to buffer families from feeling strain. In other words, good partnership quality with less conflict and higher satisfaction protects mothers from feeling financially burdened, which seems to be true independent of the family type. Furthermore, our results support the findings of Ginther and Pollak (2004), i.e., we also found that when we controlled for more than solely family type, some negative effects disappeared.

The strength of our data is certainly that we were able to merge together and examine not only information on socioeconomic issues but also subjective measures, such as financial strain and partnership quality, as well as measures on child outcomes, even those reported by children themselves. Considering all these factors together is rare and especially useful when studying family complexity. This approach helps us to understand how divorce or separation affects each child differently. Our results highlight the importance of examining not only income but also subjective financial strain, which holds especially true for controlling the effects of financial situations on different family types.

The limitations of this study are that the SDQ was only self-reported by children aged 9 to 17; therefore, those analyses are restricted to this age group. It would also be interesting to look at younger children. The sample suffers from a selection of highly educated individuals; hence, for our research question, including more people with lower education levels would have been useful because they may suffer more from financial strain. Additionally, we cannot make any inferences about causality because we have only cross-sectional data.