Parents' living arrangement and the political and civic attitudes of 13-and 14-year-old children

This paper focuses on the role of parents’ living arrangement in the attitudes of their children toward society. It uses data from the International Civic and Citizenship Education Study conducted by the International Educational Association in 2009, which surveys the civic attitudes and participation of 13-and 14-year-old students across 38 countries. In this paper, we use only 22 European countries and examine adolescents’ attitudes toward trust in civic institutions, positive attitudes toward one’s own country, equal rights for all ethnic groups, and positive attitudes toward gender equality. We distinguish between five living arrangements: two-parent families, stepfamilies, single fathers, single mothers who live alone with their children, and single mothers in multi-generational households. The analysis shows strong and significant differences between the civic attitudes of 13-and 14-year-old students living in a two-parent family and those living in other family forms, irrespective of the country, and also after controlling for parental socioeconomic background.


Zusammenfassung
In diesem Artikel wird darauf fokussiert, welchen Einfluss die elterliche Familienform auf die Einstellungen ihrer Kinder gegenüber der Gesellschaft ausübt.Datengrundlage ist die International Civic and Citizenship Education Study, die 2009 von der International Educational Association durchgeführt wurde und in der in 38 Ländern Daten zu den staatsbürgerlichen Einstellungen und staatsbürgerlichen Aktivitäten 13-und 14jähriger Schüler(innen) erhoben wurden.Wir beziehen uns in diesem Beitrag nur auf 22 Länder Europas und untersuchen die Einstellungen dieser Heranwachsenden bezüglich ihres Vertrauens in die staatsbürgerlichen Institutionen, im Hinblick auf positive Einstellungen gegenüber dem eigenen Land, zu gleichen Rechten für alle ethnischen Gruppen sowie zur Gleichheit der Geschlechter.Wir unterscheiden zwischen fünf Familienformen: Zwei-Eltern-Familien, Stieffamilien, alleinerziehende Väter, Mütter, die alleine mit ihren Kindern zusammenleben und alleinstehende Mütter, die in Mehrgenerationenhaushalten leben.Die 1 A message from the Editorial Board: Our esteemed colleague Jaap Dronkers (Maastricht University) passed away on March 30th 2016.A few weeks prior to his passing, Jaap Dronkers has resubmitted his paper after revision.In a final reviewing step, it has been suggested that some minor changes should be made before publication.The Editorial Board then decided that this article should be published in honour of our late colleague.Final revisions to this article have been carried out by Michaela Kreyenfeld, a member of the Editorial Board.We also would like to thank Lena Klein (Faculty Assistant at Hertie School of Governance) for language editing.

Introduction
Until today, sociological research on the effect of parental divorce on children has focused on measures of children's well-being or behavior, such as educational performance, psychological well-being, health, marriage, partnership, and divorce behavior.Some few studies examined the role of parental divorce in gender role attitudes (Kiecolt/Acock 1988;Wright/Young 1998;Lont/Dronkers 2004).However, there is a dearth of studies that investigated the role of parental divorce in other attitudinal measures, such as civic or political attitudes (see however Dolan 1995, or Prokic/Dronkers 2009).Even Amato's (2000) well-known review on divorce studies makes no references to the civic or political attitudes of children of divorced parents.This paper makes a modest contribution to this literature by examining how parents' living arrangements correlate with the attitudes of their children.
The data set for our investigation comes from the International Civic and Citizenship Study (ICCS) 2009, collected by the International Educational Association (IEA).It surveyed societal attitudes of students in eighth grade who were around 13 and 14 years old at time of interview.In the survey, the students were queried with whom they normally lived (father, mother, stepfather, stepmother, siblings, grandparents).Based on this information, we have generated a variable that indicated the family form at time of interview.A disadvantage of this measurement is that the ICCS 2009 data provide information on the family form at time of interview, but do not include retrospective union histories of the parents.Thus, we were unable to tell whether the parents of a respondent who lived in a single-parent household were previously married and whether the marriage had ended in a divorce or in widowhood.It can, however, be assumed that for the countries that were selected for this study the most common reason for single parenthood was divorce or separation of the parents.
Empirical studies of the relationship between parental divorce and their children's attitudes about (aspects of) society are very rare.Dolan (1995) showed that the absence of a father or a stepfather negatively affects levels of political trust in the United States.Prokic and Dronkers (2009) analyzed the societal attitudes of children in single-mother, single-father, and two-parent families in different societies with the 1999 Civic Education Study (CivEd).Although the measurement of the family form was very poor in the CivEd 1999 (it did not distinguish between stepparents and biological parents), it found some variation in the societal attitudes of children living in different family forms across different societies.Using longitudinal data from the German Socio-Economic Panel, Hener et al. (2015) found a strong relationship between growing up in a "non-intact" family and children's civic, social, and political engagement as adults.They also showed that the duration of time spent in a "non-intact family" as a child has a negative effect on participation as a young adult.Finally, Voorpostel and Coffé (2014), using Swiss longitudinal data, found that parental separation has a negative effect on young adults' voting and volunteering patterns.They partly explained this negative effect by the lower levels of political and civic engagement among separated parents compared with parents who live together.However, there are various pathways by which parental divorce affects children's attitudes about (aspects of) society.The secure attachment of children to relevant adults (among whom biological parents are paramount) is an important condition for the children's balanced psychological development.Parental divorce might affects the attachment of the children to their biological parents, due to either the divorce itself or parental conflicts before and after the divorce.Social learning of civic and societal attitudes and behavior by parental example could also influence children's own attitudes, and parental civic and societal attitudes could in turn be affected by the experience of their divorce.In addition, the children's forced choice to live with either the father or the mother after the breakup can affect their attitudes and values through the biased socialization of the coresident single parent.Finally, stressful events such as parental divorce can influence a child's long-term development, for instance by interaction with some genetic functions.This paper's aim is to contribute to the sparse literature on the influence of parental divorce and separation on adolescents' civic and political attitudes in a cross-national comparison.We do so by examining the correlation between parents' living arrangement and children's attitudes in 22 European countries.We provide separate descriptive statistics by country.However, it is beyond the scope of this paper to examine the countryspecific patterns in great detail.Another caveat of our investigation is that we do not model the causal influences of divorce and separation on children's attitudes.Instead, we provide a cross-sectional investigation that correlated parents' living arrangements and children's attitudes.We control for major confounders.However, we cannot rule out that unobserved characteristics violate our findings.

Theoretical background
The correlation between parental divorce and/or separation and children's attitudes may originate from different processes: 1) insecure attachments induced by parental conflicts and divorce, 2) social learning, 3) restructuring of parental gender roles in a single-parent family, 4) imbalanced socialization in single-father and single-mother families, and 5) interactions between stressful events and genetic function.

Attachment
Attachment theorists, starting with Bowlby (1973Bowlby ( , 1980Bowlby ( , 1999)), argue that children need a secure relationship with adult caregivers for a healthy development.Attachment theory proposes that infant behavior associated with attachment is primarily a process of proximity seeking to an identified attachment figure in stressful situations for the purpose of survival.Infants become attached to adults who are sensitive and responsive in social interactions with the infant and who remain consistent caregivers for some months during the period from about six months to two years of age.Children begin to use attachment figures (familiar people) as a secure base from which to explore.Parental responses lead to the development of patterns of attachment, which in turn lead to internal working models that will guide the child's feelings, thoughts, and expectations in later relationships.Separation anxiety or grief following serious loss are normal and natural responses for an attached infant.An extreme deficit in appropriate parenting can lead to a child's lack of attachment behavior and can result in the rare disorder known as reactive attachment disorder.Ainsworth (1967), an important figure in the formulation and development of attachment theory, introduced the concept of the secure base and developed a theory of a number of attachment patterns, or styles, in infants, in which distinct characteristics were identified: secure attachment, avoidant attachment, anxious attachment, and, later, disorganized attachment.Other theorists subsequently extended attachment theory to adults.Methods exist for the measurement of attachment patterns in older infants and adults, although measurement in middle childhood is problematic.In addition to children's care seeking, one can construct other interactions that include components of attachment behavior, including peer relationships of all ages, romantic and sexual attraction, and responses to the care needs of infants or sick or elderly adults.Although, in the early days, academic psychologists criticized attachment theory, it has become the dominant approach to understanding early social development and has given rise to a great surge of empirical research into the formation of children's close relationships (Rutter 1995).There have been significant modifications as a result of empirical research, but the main attachment concepts have become generally accepted (Bowlby/King 2004).Given that the children of divorced parents are prone to have a less secure attachment to significant others, we hypothesize that, relative to children living in two-parent families, children living in single-parent families have less trust in societal institutions and their own country and have more negative attitudes toward "outsiders" or ethnic minorities (hypothesis 1).

Social learning
Social learning theory suggests that children learn from their parents about the world, how they fit into it, and how they should behave in it (Jennings/Niemi 1968;Dalton 1980;Verba et al. 2005;Jennings et al. 2009).In addition, the parents' own behavior, including their participation in politics and community, functions as an example for their offspring (Bengtson et al. 2002).Stoker and Jennings (1995) and Voorpostel and Coffé (2012) found that parents decrease their own levels of political and civic participation following separation.Voorpostel and Coffé (2014) found lower rates of parental voting after separation, which, in turn, partly explains the lower levels of voting among children of divorce.Consequently, children living in a single-parent household learn less about civic and political participation and thus could have different societal attitudes.Thus, the conclusions from this concept are the same as from the attachment theory.

Restructuring of parental gender roles in single-parent families
According to social learning theory, children acquire sex-typed behavior by imitating significant others as role models (Stevenson/Black 1996).Children learn that mothers and fathers perform different tasks and this learning is different in single-mother and singlefather families, compared with families with both parents.Two theories suggest why this differential learning of gender roles takes place.According to role-restructuring theory, specialization by gender is more difficult in single-parent families, since single mothers and fathers must perform a wide range of tasks, including ones that are non-traditional for their gender.Hence, children in single-parent families, irrespective of the single parent's gender, should be less likely than children in two-parent families to learn traditional gender roles.If this role-restructuring theory is correct, one would expect children living in single-mother and single-father families to have less traditional attitudes about gender roles.
In contrast, according to father-absence theory, the impact of growing up in a singleparent family depends on the single parent's gender.Fathers are more likely than mothers to stress conformity to traditional gender roles.Hence, a single-father family should still instill more traditional gender attitudes in children than growing up with a single mother, with children from intact families in an intermediate position (Kiecolt/Acock 1988).Wright and Young (1998) found that, in the United States, children in father-headed families have more traditional gender-related attitudes than children in mother-headed families, supporting father-absence theory.However, the authors found gender-specific effects after controlling for maternal employment.Lont and Dronkers (2004) found that, for the Netherlands, secondary school students in single-mother families had less traditional views on future task division in upbringing, such as caring for children, cooking, other domestic chores, and earning money, than comparable students in two-parent families.But they found no difference in these attitudes regarding future task division between students in single-father families and two-parent families.
Given this overall support for father-absence theory, we hypothesize that children living in a single-mother family have more positive attitudes toward gender equality than children living in a two-parent family do.We also hypothesize that children living in a single-father family have less favorable attitudes toward women's rights and gender equality than children living in a two-parent family do (hypothesis 2).

Imbalanced socialization in father-only families and mother-only families
We can distinguish between mother-only and father-only families.At the start of the second part of the 20th century, it was customarily the mothers who got custody of the chil-dren, and the children lived mostly with their mother, while the biological father lived separately, seeing his children less or more often.The amount of time the divorced biological father spent with his children depended on the post-divorce development of relations between the ex-spouses and both ex-spouses' potential new post-divorce partnerships.As the percentage of divorced parents grew in the last part of the 20th century, it became less exceptional for the biological father to obtain custody of his children and for his children to live with him.However, until today, the majority of children live with their biological mother after divorce or separation.
This societal preference toward children staying with their mothers could cause unmeasured selectivity effects.Children living only with their father are still more exceptional than children living with only their mother.It might be that these fathers are more positively selected for the difficult task of raising children after divorce than these mothers and, consequently, the children living in father-only families are better socialized than those living in mother-only families.However, if women are, on average, more or differently equipped to raise children single-handedly than single fathers (due to either biological characteristics or cultural roles), the children of single mothers are better socialized than those of single fathers.For these two reasons, we assume that the attitudes of children in mother-only families will be different from those of children in father-only families, although the direction of these differences is not clear due to the contradictory processes outlined.
These attitudes could also differ between children in mother-only families and fatheronly families because of the gender imbalance in both types of family forms: the missing biological father in the mother-only family and the missing biological mother in the father-only family.According to role-restructuring theory, single biological mothers and fathers must perform a wide range of tasks, including ones that are non-traditional for their gender.This does not necessarily imply that these single parents do know how to balance these various gender roles in the socialization of their children, since conflicts about gender roles are themselves one of the important reasons for divorce in modern societies.As a consequence of this gender role imbalance in single-parent families, female values and attitudes will be emphasized more during socialization in mother-only families, while male values and attitudes will be emphasized more during socialization in father-only families, both in comparison with socialization in two-parent families.Murray and Sandqvist (1990) showed that children in mother-only families have relatively worse math grades compared with their reading grades, while children in father-only families have relatively worse reading grades compared with their math grades.
Given these differential emphases of male and female values and attitudes during socialization in father-only and mother-only families, we hypothesize that children in mother-only or father-only families will have different attitudes, reflecting the dominant gender roles in these single-parent families.The traditional female gender role is more related with activities of caring for and nurturing children and those who need help, while the traditional male gender role is more related with defending territory against intruders and competition.As far as these traditional gender roles still influence socialization in singleparent families, we expect more positive attitudes toward ethnic groups and gender equality in mother-only families, while children in father-only families will have more positive attitudes toward civic institutions and one's country (hypothesis 3).

Interactions between stressful events and genetic function
It is well known that stressful events such as parental divorce and death can influence a child's long-term development.Most previous studies have focused on how a child's individual characteristics and genetics interact with that particular child's experiences in an effort to understand how health problems emerge.Romens et al. (2014) were recently able to measure the degree to which genes were turned "on or off" through a biochemical process called methylation.They found an association between the kind of parenting children had and a particular gene (the glucocorticoid receptor gene) responsible for crucial aspects of social functioning and health.Not all genes are active at all times.DNA methylation is one of several biochemical mechanisms that cells use to control whether genes are turned on or off.Romens et al. (2014) examined DNA methylation in the blood of 56 children aged 11-14 years, half of whom had been physically abused.The researchers found that, compared to children who had not been maltreated, the abused children had increased methylation at several sites of the glucocorticoid receptor gene, also known as NR3C1, echoing the findings of earlier studies of rodents.In their study, the effect occurred on the section of the gene that is critical for nerve growth and an important part of healthy brain development; however, there were no differences in the genes with which the children were born.Instead, the differences were seen in the extent to which the genes had been turned on or off.The gene identified by the researchers affects the hypothalamic-pituitary-adrenal axis in rodents.Disruptions of this system in the brain make it difficult for people to regulate their emotional behavior and stress levels.Circulating through the body in the blood, this gene affects the immune system, leaving individuals less able to fight off germs and more vulnerable to illness.Given that parental divorce or separation is also a stressful event, genetic factors could influence a child's long-term development and thus the child's attitudes toward society.This also leads to the assumption that children of separated parents have less trust in societal institutions and adopt more negative attitudes toward outsiders in their society.This is very much in line with hypothesis 1.Unfortunately, we are unable to isolate the role of genetic and socialization factors in our analysis.

Differences between male and female students
Civic attitudes might have different levels of relevance for male and female students (just as skills in reading and math are still unequally distributed among boys and girls, even in the most gender-equal societies).In addition, the restructuring of parental gender roles and possible imbalance between gendered values and attitudes in mother-only and fatheronly families could have different consequences for male and female students.We will therefore analyze civic attitudes separately for male and female students.

ICCS
This paper uses the International Civic and Citizenship Education Study (ICCS) data obtained by the IEA in 2009.This survey focuses on the extent to which young people are ready to take on their role as citizens in democracies.It collects information on the civic attitudes and participation of 13-and 14-year-old students across 38 countries across the world (Brese et al. 2011).In this paper, we use only European countries.
The ICCS 2009 student target population was students in the grade that represents eight years of schooling counted from International Standard Classification of Education (ISCED) Level 1, provided that the average age of students in this grade was 13.5 years or above at the time of the assessment (usually Grade 8).To obtain accurate and representative samples, ICCS 2009 used a two-stage sampling procedure whereby a random sample of schools per participating country is selected at the first stage and one or two intact target grade classes is sampled at the second stage.Students of the target classes had to answer a written questionnaire.All students of the targeted grade classes had to fill in the questionnaire, irrespective of their age.Relatively old students have repeated one or more classes, while relatively young students might have skipped classes.
Although the ICCS focuses particularly on the role of schooling in the development of civic attitudes and participation, it is limited in the available background variables.It does not contain background variables such as family income and wealth or union histories of the parents.Available background variables are father's and/or mother's educational level and occupational status.ICCS 2009 combined these four indicators in a national index of socioeconomic background of the students.This index is normalized (national average zero) and standardized.
For the purpose of this study, we use the measurements of civic attitudes of 13-and 14-year-old native students in 22 European countries: Austria, Flemish Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, England, Estonia, Finland, Greece, Ireland, Italy, Latvia, Lithuania, Luxembourg, the Netherlands, Norway, Poland, Slovakia, Slovenia, Sweden, and Switzerland.In total, we analyze 59,545 native children in 22 different European countries.We thus exclude the population with a migration background.

Dependent variables
The ICCS measures various aspects of civic attitudes.In this paper, we use the following indicators, which are scales of various items: -Trust in civic institutions.Scale based upon trust in six institutions: national government; local government of your town or city; courts of justice; police; political parties; national parliament.-Positive attitudes toward one's own country.Scale based upon seven items: flag of my country is important to me; political system in my country works well; I have great respect for my country; in my country, we should be proud of what we have achieved; I am proud to live in my country; my country shows a lot of respect for the environment; generally speaking, my country is a better country to live in than most other countries.-Equal rights for all ethnic groups.Scale based upon five items: all ethnic/racial groups should have an equal chance to get a good education in our country; all ethnic/racial groups should have an equal chance to get good jobs in our country; schools should teach students to respect members of all racial/ethnic groups; members of all ethnic/racial groups should be encouraged to run in elections for political office; members of all ethnic/racial groups should have the same rights and responsibilities.-Positive attitudes toward gender equality.Scale based upon six items: men and women should have equal opportunities to take part in government; men and women should have the same rights in every way; women should stay out of politics; when there are not many jobs available, men should have more right to a job than women; men and women should get equal pay when they are doing the same jobs; men are better qualified to be political leaders than women.
ICCS computed the weighted likelihood estimation scores for these scales in such a way that the mean is 50 and the standard deviation is 10 for equally weighted countries.We use these scores in this paper.The average scores for these scales and the prevalence of various family forms vary substantially between the 22 European countries.We do not focus on this cross-national variation; however, we add country dummies into the regression to account for differences across countries (see Tables A1-A2 in the Appendix for the descriptive statistics broken down by country and Table A3 for the total sample).

Family forms
A disadvantage of the ICCS 2009 is that it lacks the complete union history of the respondents' parents.Thus, we rely on information on the living arrangement at time of interview as a proxy for whether the respondent experienced parental separation or divorce.
Although we assume that, in most European countries, divorce or separation is the most common reason for single parenthood, there could be other reasons for growing up in a single-parent family (with or without a guardian) than separation or divorce, such as death of one of the parents.In order to find an operational definition of family forms, we rely on the students' responses, who were asked to report with whom they regularly live.Thus, the living arrangements as perceived by the students, and not by parents or authorities, are used to classify peoples' living arrangements.Separation and divorce are treated alike, as separation after cohabitation has more or less the same effect on children compared to divorce after marriage (Härkönen/Dronkers 2006;Dronkers/Härkönen 2008).Married parents who stopped living together but did not get a formal divorce yet, are also treated the same way as formally divorced parents.This aspect is especially relevant for Catholic countries such as Italy, Ireland, Portugal, and Spain, where a formal divorce is still difficult to obtain.A disadvantage of this measurement is, however, that some children may live without a parent temporarily (e.g., children of fishermen and of fathers working in foreign countries) which we may misclassify.
The students could indicate the following as living at their home: the mother, a female guardian, the father, a male guardian, siblings, grandparents, and others.We combined these answers into eleven different family forms.We deleted all family forms with less than 500 cases, which left us with the following categories: -Two-parent family: Mother and father, irrespective of whether there were also grandparents or others living in the home (N=47,837) -Single mother: Only the mother without a male guardian or the grandparents, irrespective of whether there were others living in the home (N=6,692) -Stepfather family: The mother and a male guardian, irrespective of whether there were also grandparents or others living in the home (N=3,732) -Single father: Only the father without a female guardian or the grandparents, irrespective of whether there were others living in the home (N=760) -Mother and grandparents: Grandparents with a mother but without a father, irrespective of whether there were others living in the home (N=1,538).

Control variables
We use as controls in the regression analyses: a national index of socioeconomic background (as constructed by the ICCS 2009 and based upon the highest educational levels and occupational status of fathers and/or mothers), the literacy of the home of the students (based on the number of books), and gender.We also conduct "sensitivity checks" where we included age of the student and years of expected further education.The problem with these two additional control variables is that they are also indicators of possible consequences of family form.A relative high age of the student might indicate a repetition of classes (and thus lower educational performance in the past), and this might be related with consequences of divorce and separation.Years of expected further education might also be influenced by the experience of parental divorce and separation.Thus inclusion of these two control variables might lead to 'over-control' of the relation between civic attitude and family form.

Results
The top panel of Table 1 shows the differences in societal attitudes between students living with both parents and students living in other family forms, with no controls for parental socioeconomic background or gender but with controls for the 22 European countries of our study.Students living in single-mother families and stepfather families scored significantly lower on all four societal attitudes, compared to students living with both parents.Students living in mother-grandparent families scored significantly lower on only two societal attitudes (trust in civic institutions, gender equality).Students living in single-father families scored significantly lower on two societal attitudes (trust in civic institutions, gender equality) and significantly higher on the two other (positive attitude toward one's own country, equal rights for ethnic groups).
The bottom panel of Table 1 shows the differences in societal attitudes between students living with both parents and students living in other family forms, with additional con-trols for parental socioeconomic background and gender.Students living in stepfather families scored significantly lower on all societal attitudes, compared to students living with both parents.Students living in single-mother families scored significantly lower on two societal attitudes (trust in civic institutions, gender equality).Students living in single-father families scored significantly lower on one indicator of societal attitudes (trust in civic institutions) and significantly higher on two others (positive attitude toward one's own country, equal rights for ethnic groups).Students living in mother-grandparent families scored not significantly different than students living with both parents.The inclusion of age and expected years of student education does not change the overall pattern.Source: Author's weighted computations from ICCS 2009 data.The parameters for the country dummies that are included in all the equations are not shown.
Table 2 shows an interaction of country and family forms.Due to sample size problems, we only compared children living with single mothers and in stepfamilies and compared them to children who live with both parents.By computing these differences by country, we lose statistical power and thus many differences are no longer significant.Students in single mother families in England, Finland, Italy, the Netherlands, and Sweden have less trust in civic institutions.Students in single mother families in Austria, Denmark, England, Finland, Italy, the Netherlands, Norway, and Sweden have a less positive attitude toward their own country.Results for the other two societal attitudes per country are less pronounced.
Table 2: Societal attitudes of 13-and 14-year-old native students in single mother and stepfather families compared to two-parent families (reference group).OLS regression standardized coefficients.

Gender equality Single mother
Stepfather family

Single mother
Stepfather family

Single mother
Stepfather family

Single mother
Stepfather family

Austria
Table 3 shows the differences in societal attitudes between male and female students living with both parents and students living in other family forms, controlling for parental socioeconomic background, gender, and countries.The main outcome of this table is that there are only very minor gender-related differences in societal attitudes.We, thus, need to refute the hypothesis that separation and divorce affect female and male children differently.Source: Author's weighted computations from ICCS 2009 data.The parameters for the country dummies that are included in all the equations are not shown.We controlled for socioeconomic background and home literacy.

Conclusion
In this paper, we analyzed the relation between parents' living arrangement and the political and civic attitudes of their children.Data came from the ICCS which collected attitudes of 13-and 14-year-old students.Albeit that the data included non-European countries as well, we limited our investigation to the 22 European countries that were available in this data.We used four attitudes toward society: trust in civic institutions, positive attitudes toward one's own country, equal rights for all ethnic groups, and positive attitudes toward gender equality.The main independent variable was the living arrangement of the parents.Here we distinguished: two-parent families, stepfamilies, single fathers, single mothers who live alone with their children, and single mothers in multi-generational households.A disadvantage of the ICCS 2009 is that it lacks information on the union histories of the parents.Although we assumed that, in most European countries, divorce or separation is the most common reason for the single parenthood of parents of 13-or 14-year-old children, there could be other reasons for growing up in a single-parent or stepfamily, in particular the death of a parent.Another disadvantage of ICCS 2009 is that it contains not many background variables.The available socioeconomic index is a combination of the occupational status and the educational level of both parents and home literacy measures the number of books at home.However, including more control variables (like age of student and expected years of further education) did not substantially change the differences in societal attitudes of children living in different family forms.
Our first hypothesis was that, relative to children living in two-parent families, children living in single-parent families have less trust in societal institutions and their own country and have more negative attitudes toward "outsiders" and ethnic minorities.This assumption was only partially supported.While we found that children in single-parent families express fewer trust in political institutions than children in "intact" families, we do not find significant differences in the attitudes toward ethnic minorities.
Our second hypothesis stated that children living in a single-mother family have more positive attitudes toward gender equality than children living in a two-parent family, while children living in a single-father family have less favorable attitudes toward gender equality.This hypothesis is not supported by our data.
Our third hypothesis stated that more positive attitudes toward ethnic groups and gender equality would be found in mother-only families, while children in father-only lies would have more positive attitudes toward civic institutions and one's country.This hypothesis is only supported for more positive attitudes toward one's country of fatheronly families.
Overall, our investigations support the findings from other European studies on the relations between civic and political attitudes and family form (Prokic/Dronkers 2009;Hener et al. 2015;Voorpostel/Coffé 2014).Of course, our analysis is just a small step along a long road, but it suggests that parental separation affects more than just the wellbeing of children.Given that divorce and separation are the most common reasons for the single parenthood of parents, our results suggest that divorce and separation have consequences outside the family, in this case the amount and the direction of societal attitudes by young members of society.
At the same time, it is important to emphasize that the found variation of societal attitudes of children living in different family forms does not necessarily mean that the family form must be the cause of these differences.We could only control for some background variables, while we do not have a pre-and post-divorce/separation measurement of societal attitudes.

Table 1 :
Societal attitudes of 13-and 14-year-old native students in different family forms (with the two-parent family as the reference group).OLS regression standardized coefficients and t-values in parentheses.

Table 3 :
Societal attitudes of 13-and 14-year-old male and female native students in 22 different European countries: Deviations of the scores of students in two-parent families.OLS regression (standardized coefficients)