1 Introduction

Many researchers have provided definitions of children’s subjective well-being. All of these researchers agree that subjective well-being refers to different definitions of the state of mind, such as satisfaction with life (Diener 1984; Casas and Rees 2015; Diener and Biswas-Diener 2002; Rees et al. 2016). Satisfaction is a state of mind linked to many relationship dimensions (family, friendship, work, school, neighbors). The difficulty lies in finding an appropriate indicator that roughly measures the complexity of this state of mind, which is even more complicated in the case of children due to their difficulty in understanding this phenomenon. Despite this difficulty, we find numerous research papers, especially from the field of psychology, that have created indices of indicators to measure children’s subjective well-being from the available data (Diener 1984; Huebner 1991; Huebner et al. 2004; Cummins and Gullone 2000; Casas et al. 2013; Russell 2003; Cummins and Lau 2005). In this paper, we have used the items from Russell’s (2003) scale on core affects (CAS) in a short version of 6 items (satisfied, happy, relaxed, active, calm, full of energy) because we consider that it is the one that best minimizes cultural differences, allowing us to carry out the comparative analysis that we propose. Russell (1980) suggests that persons possess a cognitive conceptual structure capable of representing affect summarized by a simple circumplex model of affect. According to Leung and Bond (1989), this theoretical model can be used to guide the selection of antecedent variables to minimize the impact of cultural biases. This model captures the cognitive process structure of emotions summarized in a scale by a simple circumplex model that would be reproduced in cross-cultural countries (Russell 1980). This scale facilitates the measurement of the universal conditions for children’s well-being by international comparative studies.

These indices try to basically measure states of mind related to subjective well-being in order to associate the state of mind with the behaviors and personality aspects of children and young people. However, there are few studies that focus on analyzing how family structures and relationships, as well as socioeconomic conditions, affect children’s subjective well-being. These studies suggest that socioeconomic status and family structure have limited effects on satisfaction with life (Vandewater and Lansford 1998; Holder and Coleman 2009). However, Klocke et al. (2013) use multilevel analysis and find that family structure and school interactions explain the differences in children’s subjective well-being across countries.

Therefore, the literature reflects the controversy about the effects of family structure and family dynamics on SWB (Schoppe et al. 2001; Härkönen et al. 2017). However, the importance of family relationships, friends and school interactions for children’s well-being and social-emotional development is widely recognized in the literature (Gilman and Huebner 2003; Nickerson and Nagle 2004; Bornstein 2007; Lee and Yoo 2015).

In regard to the family, parents invest economic and emotional resources to achieve the well-being of their children. The former are obtained through work and the latter through time spent with the children and the quality of the relationship with them, which in turn depend on the type of family in which the children live and the quality of the relationship between the parents (Thomson et al. 1994; Berger and McLanahan 2015). There is extensive literature documenting positive associations between family income and children’s health and development (Duncan et al. 2014). In addition, a large body of research documents that the quality of parenting is associated with a positive cognitive and social-emotional outcome for children (Steinberg 2001; Schoppe et al. 2007a). Research on parenting has developed in recent years, due in part to empirical demonstrations that the quality of the parental relationship correlates with children’s outcomes (McHale et al. 2002; McHale 2007). However, there are few studies on how children’s well-being varies according to the type of family in which the child lives. Most studies show that children who live with married biological parents have higher well-being than children living in other types of family (single-parent families, cohabitant families or step-families) (Manning and Lamb 2003; Manning and Brown 2006; Artis 2007; Hofferth and Casper 2007; Klausli and Owen 2009; Brown 2010). The studies carried out on the concept of “fragile families and child well-being” (Waldfogel et al. 2010; Berger and McLanahan 2015) refer to how the type of family in which the children live affects their personality and behavior, but they do not associate the family structure with children’s subjective well-being, and this is one of the novel contributions of our analysis. According to the findings of Berger and McLanahan (2015), children living with both their biological parents show greater cognitive skills and fewer behavioral problems than children in other types of family. However, the researchers found that married biological fathers show lower levels of engagement with their children than cohabiting biological fathers and married and cohabiting stepfathers, whereas maternal engagement with children does not differ by family type. Although these results are inconsistent with the findings in the majority of previous empirical studies, they are consistent with the findings from several recent analyzes of FFCW data (Gibson-Davis 2008; Carlson and Berger 2013) about the effects of family type on behavior and cognitive development. While from these studies one can hypothetically argue that children’s subjective well-being is greater for children living in so-called “intact” families, there is in fact no empirical evidence to corroborate this possible association.

There is considerable evidence in studies about the connection between the parent–child relationship quality and child outcomes (externalizing problems and cognitive development) (Tamis-LeMonda et al. 2004; Cabrera et al. 2011; Schoppe et al. 2007b; Cummings and Merrilees 2014; Berger and McLanahan 2015). However, there are few studies that analyze the relationship between the quality of the interaction between parents and their children and SWB (Cowan et al. 2010).

In fact, there is extensive literature that shows that satisfactory family relationships, relationships with friends and a good relationship with teachers, the school and the neighborhood contribute to children’s subjective well-being even in different cultural contexts and in economically disadvantaged situations (Holder and Coleman 2009; Lee and Yoo 2015). These findings suggest that children’s subjective well-being is affected more by micro factors, such as the family, friends and school, than by macro societal factors.

Although there have been numerous studies on the well-being of children, few studies have looked at how these associations vary in different countries and how they relate to other factors, such as family structure, relationships with families, friends and the school. This is largely due to the methodological complexity involved in interviewing children (Oishi et al. 1999; Park and Huebner 2005; Rees et al. 2010; Casas 2016). In fact, existing studies have focused on analyzing only how children’s subjective welfare varies across countries in relation to variables such as poverty, health, education and housing. Bradshaw et al. (2013) compared children’s subjective well-being across European nations, and they found an association between subjective well-being and all the other objective domains at the country level, such as material, health, education, behavior and housing environment. Klocke et al. (2013) used multilevel analysis to examine the effects of country-level differences, such as economic growth and youth unemployment, on children’s subjective well-being, while controlling for variations at the individual and school levels. They concluded that the macro (country) variables do not explain how children’s subjective well-being varies across countries. However, they found that the micro (family) and meso (school) levels explain the variation. According to Lee and Yoo (2015), the frequency of family activities, frequency of peer activities and neighborhood safety are the factors most related to the levels of children’s subjective well-being across countries. The findings of Holder and Coleman (2009) suggests that multiple dimensions of social relations (family, friends, teachers), including both positive and negative interactions, are associated with the happiness of children. Thus, relationships, talking to parents and peers and playing with friends have positive effects on subjective well-being, while situations such as bulling have negative effects (Nickerson and Nagle 2004; Huebner et al. 2006; Smith 2013). Studies also show that the neighborhood and safety are relevant factors affecting children’s subjective well-being (Coulton and Korbin 2007; McDonell 2007; Lee and Yoo 2015).

1.1 Study Aims

The final purpose of this study is to provide new empirical evidence on the possible association of children’s wellbeing with factors such as family structure, relationship of children with their families, friends and school. It is a issue practically unexplored in the scientific literature for children between 9 and 12 years old. This analysis incorporates the comparative perspective through data collected in Germany, the United Kingdom, Norway and Spain, based on the possibilities offered by the variables included in the International Survey of Children’s well-being. In this paper we investigate to what extent, cross-country and institutional variations exist in regard to variables such as, family relationship, friends, school considered as predictors of SWB. The novelty of this study is the presentation of new empirical findings on the factors related to the subjective well-being of children aged below 12 across different institutional contexts for children under 12 years of age.

We also expect that the associations of family income (occupational situation), parental relationship with children, social relationship with friends, school and community will contribute to children’s subjective well-being considered as a complex process of the state of mind interacting with numerous relational factors.

2 Methodology

2.1 Participants: Data Source and Sample

We used data from the International Survey of Children’s well-being (ISCWeB). The ISCWeB is a worldwide research project on children’s subjective well-being. Data collection for the pilot study took place between winter 2011 and winter 2012 in 14 different countries in two age groups (10 and 12 years old). The ages of the children were grouped together (10–12 years) because according to the available evidence, the observable differences are minimal (Rees and Main 2015). The criterion for the grouping was belonging to the group of students attending the second cycle of primary education, with slight variations by country. Due to the sample size, it was decided to group participants by ages, although minimal differences could be expected by sex in each age group. The sample was based on schools in all of the countries. Since the survey was carried out as a pilot study, the convenience sampling methodology was employed in most of the countries.

We have considered only 4 countries (Spain, Germany, United Kingdom, Norway), as these are representative of different welfare state models with very different family and childhood policies (Esping-Andersen 1999). This selection of countries is based on the idea that children’s subjective well-being is not entirely connected to factors but it is related to the country’s policies, environment, and culture (Lee and Yoo 2017). In the existing literature on child well-being, several studies have motivated this grouping and selection of countries (UNICEF 2013; Rees and Main 2015; Lee and Yoo 2017; Yoo and Ahn 2017). On the other hand, there is fruitful literature using the welfare model to identify groups of countries with similar profiles in terms of diverse sociocultural features, such as child poverty and social exclusion (Labato and Kaup 2014), family policy (Gauthier 2002), quality of life (Dorothy et al. 2014), and culture heterogeneity (Eid et al. 2003) The sample was composed of 4812 pupils, range in age between 10 and 12, of which 49.2% are boys and 50.5% girls.

2.2 Variables

The key dependent variable in the study is the measure of subjective well-being, using Russell (2003) core affect scale. The dependent variable was constructed from the questionnaire: “Below is a list of words that describes different feelings and emotions. Please read each word and then tick a box to say how much you have felt this way during the last two weeks: (Satisfied, happy, relaxed, active, calm, full of energy:) (not at all (0); Extremely (10))” to which the scale of Russell’s core affect was applied. In this work, we used the items from Russell’s core affect scale (Russell 2003) in a short version of 6 items of 11 items (satisfied, happy, relaxed, active, calm, full of energy), calculating the mean of all items. Some researchers recommend transforming the psychometric scales of subjective well-being into a 0–100 scale to facilitate comparison (Cummins and Gullone 2000). In this case, an 11-point scale, from “not at all” to “extremely”, was used. Therefore, Russell’s core affect scale, our dependent variable, relates to different dimensions of subjective well-being (feeling happy and satisfied). The dependent variable was constructed by adding the 11 items of Russell’s scale and applying the arithmetic mean (Table 1).

Table 1 Variables
Full size table

As shown in Table 1, the independent variables include the family, school and community factors that were related to children’s subjective well-being in previous studies. The independent variables were constructed from the responses that children of 10 and 12 years old gave to the respective questionnaires of the survey. We used the combined data sets across age groups for compatible variables. These include all variables that were asked in the same format in each questionnaire. In the data sets combining age groups, there are also weighting variables (equal age and country weight). Each country has 1000 weighted cases in each group. Children respond to the “general” questions without considering time periods (last year, last 6 months, etc.). The questionnaire does not contemplate time periods because children may not be able to remember them.

  • Family structure was coded into in 4 categories: (two-parent family, mother-only family, father-only family and other).

  • Family relationship was coded using two indices that differentiate between the attention given to children (consideration of parents: parents who listen, good family time together, parents who treat them fairly) and time dedicated by parents (frequency of family talks, frequency of family fun, frequency of family learning).

  • Relationship with friends was measured by the time dedicated to friends (frequency of talking to friends, frequency of having fun with friends, frequency of studying with friends).

  • School life was measured by the relationships that children have with their teachers and the school in general (teachers who listen, liking school, fair teachers and a safe school).

  • Exclusion was measured by variables regarding bullying (frequency of peers being hit and frequency of peers being excluded).

  • Community was measured as neighborhood/social environment (safe areas to play and safe areas to walk).

  • Family economy was measured using two variables: worrying about family money and the number of adults in a paid job.

  • Leisure was measured as free time (frequency of classes, frequency of organized leisure, frequency of sports and exercise)

  • Gender was treated as an independent variable because gender differences have been reported in various subjective well-being studies as a key variable.

2.3 Analytical Strategy

First, we analyzed the significant correlations between the independent variables and Russell’s core affect well-being scale (as a dependent variable) for all the countries. Next, an ANOVA was carried out to examine the significance of the correlates of children’s subjective well-being separately for each country. Finally, we applied a multiple linear regression analysis to examine the relationship of the independent variables with children’s well-being for the sample as a whole and for each country. These analyses assessed how differences in family structure, economic situation, children’s perceptions of family relationships, relationships with friends, life, school, community and free time contributed to differences in children’s subjective well-being across countries. We estimated these regressions separately for each country and tested the statistical equivalence of the coefficients for each country across equations.

For the descriptive statistical analysis of the sample, the number of cases present in each category and the corresponding percentage or, depending on the type of variable, the mean and standard deviation, were obtained.

The correlation between variables was studied using the Pearson correlation coefficient (r). The comparison between groups was carried out by means of the ANOVA test after checking the normality assumptions (Kolmogorov–Smirnov test) and homogeneity of variance (Levene test), making Tukey’s pairwise mean comparisons in the cases where the ANOVA was found to be significant. A multiple linear regression model was used to determine the relationship of well-being with the variables for sex, coexistence, family, friends, school, free time and finances. The statistical analysis was performed using the SPSS 23.0 program for Windows. Differences with p < .05 were considered statistically significant.

3 Results

Table 2 shows the value of Russell’s core affect scale included in the survey by country. The findings show that children’s subjective well-being varies across the countries. Norway has the highest level of well-being, while Germany has the lowest. For overall SWB (Russel’s core), Germany has the lowest score. No significant differences were observed between Spain, the United Kingdom and Norway, but they were observed between Spain and Germany, with children’s well-being being higher in Spain.

Table 2 Children’s subjective well-being index
Full size table

First, descriptive statistics are presented for all the variables used in the model as well as the correlations between the measures of the selected variables (family relationships, relationships with friends, school, leisure time, economic concerns) and Russell’s core well-being scale. The results show the correlations that are statistically significant and positive (Table 3). Despite the interpretative limitations of these indices since they are variables that are difficult to objectify through direct research, they provide useful information for new research with similar objectives. It should be noted that all variables considered have a positive relationship with children’s subjective well-being, except for that referring to concern about the economic situation and the bullying situation in school. To go beyond the strict association between variables and to be able to estimate the possible interaction of variables such as sex, type of family in which the child lives, perception of relationships with family and friends, relationship with the school and teachers, perception of leisure time and safety of the neighborhood in which they live, we carried out an analysis of multiple linear regression models for the total sample and for Germany, the United Kingdom, Spain and Norway individually.

Table 3 Averages, standard deviations and correlations of the variables
Full size table

Table 4 shows the results of the multiple linear regression models performed for the total sample. First, we observed gender differences in SWB. Girls have lower subjective well-being than boys in all the countries. In relation to cohabiting, the children who live with both parents have higher well-being than those who live in other family types. The categories of the time devoted to them by family and friends, as well as the time spent on leisure are associated with differences in the well-being of children, so the more time children spend with parents, with friends or in leisure activities, the higher their well-being. Fair treatment by parents and teachers shows the same positive association with child well-being. As expected, greater the exclusion and bullying by schoolmates, was associated with lower the well-being. With regard to economic variables, children who show greater concern for the family finances have a lower well-being, whereas in households with a higher number of working adults, well-being is higher.

Table 4 Multiple regression analysis on the total welfare of the sample
Full size table

We can see the relationship of these variables by country in Table 5. In general, the variables maintain the same significant association with the well-being of children as they do in the overall sample, although there are some differences that should be highlighted. For example, the indicator referring to time devoted to friends is not significant in the case of Germany, unlike in the other countries considered. On the other hand, it is also worth mentioning that while bullying and exclusion by peers is seen as a factor negatively associated with child well-being in all countries, in the case of Norway, this factor is not significant. This could indicate that in Norway, this type of bullying situation in school is less frequent, as shown in the following table.

Table 5 Multiple regression analysis on well-being by country
Full size table

For a more in depth look at the comparative analysis between countries, we carried out an ANOVA test of the different variables included in the model. This ANOVA is a continuation of the previous analyses carried out by MANOVA (see Table 6). The MANOVA analysis was carried out to compare the scores of the variables between the countries considered. The calculation evidenced the existence of statistically significant differences [Wilk’s Lambda (24; 9493.3) = 20.303, p < .001; eta2 = .047] between countries. The variables included in this model were tested to compare differences between countries. As shown in Table 6, several differences were observed. Regarding children’s concerns about money, differences emerge across all the countries and more specifically Spanish children are most concerned about the economic situation. When we analyze children’s perception about attention from their parents, no differences are detected between Spain and the United Kingdom, but there are differences between Spain, Germany and Norway, with the Norwegians dedicating the most time to their children. For example, in the time parents spend with their children, there are no significant differences between Spain and Germany (they share letter a) or between Germany and the United Kingdom (they share b), but there are significant differences between Norway and the rest of the countries (do not share any letter). In Norway, parents spend more time with their children, which contributes to the children’s well-being. These compared differences are evidenced by the calculated Tukey’s pairwise mean comparisons.Footnote 1 These findings could be interpreted to indicate that in Spain and the United Kingdom, children perceive that they receive less attention from their parents than do children in Germany and Norway, mainly due to the work-family balance policies, which are less developed in Spain and the United Kingdom than in Norway and Germany. As regards time spent with friends, there are no differences between Spain and Norway, but there are differences with the rest of the countries, with Spain and Norway scoring highest in the time devoted to friends and Germany the lowest. For neighborhood, the case of Norway stands out because it has the highest values for feeling safe in the play environment. A very interesting result is that related to the relationship with teachers and the school, with clear differences between all the countries. The case of Norway stands out significantly, having higher scores for children’s perceptions of their relationships with and attention from the school and teachers. Finally, regarding bullying and exclusion by peers, Spain, Germany and Norway are similar, differing from the United Kingdom, where a higher score is observed for this variable. In contrast, Norway has the lowest value, and it therefore follows that bullying and exclusion at school occur less frequently.

Table 6 Comparison of the index of well-being and the variables related to family context, friends, school, free time and economics (ANOVA)
Full size table

In short, comparing how the independent variables are related to subjective well-being across countries shows some interesting findings. Boys reported higher levels of subjective well-being than girls in all countries, as documented in previous studies. Children living in two-parent families show higher levels of subjective well-being than those living in a mother-only or father-only family. There are no differences between countries. On the other hand, variables such as the frequency of time spent with parents and friends, the support of teachers and the time dedicated to leisure contribute to greater subjective well-being, with nuances between countries, as in the case of Germany, where relationships with friends do not seem to be as positive as in other countries. Bullying at school and concern about the family’s economic situation contribute to lower well-being in all countries. In the case of Spain, the negative relationship with the economic situation is especially relevant due to the effects of the economic crisis. In the case of Norway, the possible low incidence of bullying seems to reduce the negative effects on children’s well-being, unlike in other countries.

In short, although the differences in some of the statistical indices are not very large, they are statistically significant. In addition, from the sociological perspective, these differences suppose a great contribution to social theory. These results also have relevant implications for social policy.

4 Discussion

In this section, we will present the conclusions from this analysis of sets of different items selected to measure children’s subjective well-being in several countries. First, our analysis suggests that, with some limitations, all variables measured in the study presented acceptable validation for within-country analysis in all countries selected for the analysis.

Like previous studies, our analysis highlights that children’s subjective well-being is related with multiple factors, such as demographic factors and contextual factors (family, school, friends, social environment, material circumstances or leisure). The results emphasize that children’s subjective well-being varies according to the cultural and institutional context of each country.

First, one variable related to SWB in our findings is gender differences. Boys tend to have higher levels of SWB than girls in all the countries analyzed. These results confirm that sex has a significant role, such that girls report a lower well-being than boys, a result that has been found in numerous previous studies (Casas et al. 2013; Rees and Main 2015; Dinisman and Ben-Arieh 2016). According to Kaye-Tzadok et al. (2017), the association between gender and SWB is complex, and different factors, such as biology and culture, interact to produce complex findings. These results, found in previous research (Klocke et al. 2013; Nordlander and Stensöta 2014; Lee and Yoo 2015; Montserrat et al. 2015; Kaye-Tzadok et al. 2017), point to the need to consider gender as a social construct that is strongly connected with social relations, inequality and differential well-being from childhood (Gilligan 1982; Baker-Miller 1986). As suggested by Kaye-Tzadok et al. (2017), the observation of these gender differences in childhood is a key indicator for investigating the causes of such disparities and designing public policies that help neutralize the inequalities that have their origin in family and school socialization. Reflecting on these findings, comparing different countries and drawing on cultural and national factors, such as the level of gender equality and the welfare system within the countries, would be a valuable contribution to understanding these differential findings.

In keeping with previous studies, our findings show that, on average, children living with two parents have a higher well-being than children in single-parent families (Manning and Lamb 2003; Hofferth 2006; Artis 2007; Sawhill 2014). This factor can be related to the economic situation and the parental styles of fragile families (children living in single and cohabiting parent families (McLanahan and Beck 2010; Berger and McLanahan 2015; Rees 2017). In this respect, the literature emphasizes that in the “fragile family”,Footnote 2 there are more likely to be circumstances of economic vulnerability, lower educational levels of the parents and therefore less ability to develop parenting styles that contribute to improving children’s subjective well-being and to reducing inequality among families (Waldfogel et al. 2010; Sawhill 2014). The findings of Waldfogel et al. (2010) are consistent with those of previous research suggesting that children in fragile families are at risk of poorer school achievements.

The findings indicate that children who show greater concern for the economic situation of the family have a lower well-being, whereas in households where the number of working parents is higher, well-being increases. These findings provide new evidence since the relationship between the socioeconomic context and children’s SWB is not sufficiently clear (Bradshaw et al. 2011; UNICEF 2013).

There is extensive evidence that high-quality parental relationships and parenting behavior protect children’s well-being, particularly their social and emotional well-being. The literature has generally shown that family income, high-quality parental relationships and high-quality parenting are mediators of the association between family type and children’s well-being (Marsiglio 2004; Hofferth and Casper 2007; Klausli and Owen 2009; Carlson and Magnuson 2011). However, the findings of Berger and McLanahan (2015:15) show few statistically significant differences by family type when associating family income, parental relationship quality and parenting quality with children’s cognitive skills and behavioral problems. In the same area, the results of Dinisman et al. (2017) suggest that there are no differences in children’s well-being for children living in single-parent families and children living in care. However, the present study shows that at least in this sample, children living in two-parent families reported better SWB than those living in single-parent families and children in care. According to Berger and McLanahan (2015), differences in levels of family income, parental relationship quality and parenting quality between family types appear to operate in more complex ways (Gibson-Davis 2008; Carlson and Berger 2013; Brown et al. 2015). In our analysis, the results provide support to the literature suggesting that across countries, living with two parents has a positive relationship with children’s subjective well-being compared to living with only the father or the mother. These findings show that children’s perception of economic situation of single parent families have a negative association with children’s wellbeing. In line with the previous literature, living in a fragile single-parent family could be associated with greater economic vulnerability and less involvement of the parent in the education and care of the children. However, this positive role of the family structure is less evident when we disaggregate the analysis by country. This analysis is limited in that the survey does not provide us with data on the economic, employment or educational situation of the parents. Such data would allow us to carry out a more comprehensive study of parental styles, family structures and children’s subjective well-being. These results can also be applied to different models of welfare states with different public policies aimed at families and children, where the welfare states of northern Europe have the greatest capacity to produce well-being (Nordlander and Stensöta 2014).

We find that parental involvement, time spent with friends, leisure time and satisfying relationships with the school and the teachers contribute significantly to children’s well-being. Therefore, personal relationships (friends, school, family) are the strongest factors for happiness according to the SWB literature (Myers and Diener 1995; Diener and Oishi 2005; Lyubomirsky et al. 2006; Holder and Coleman 2009). Previous studies have verified these findings from a comparative perspective, using individual level data (Klocke et al. 2013; Lee and Yoo 2015). However, these studies did not analyze the differences existing between European countries with different cultures and different welfare systems. Our results data allowed us to identify slight variations in SWB across countries. In fact, relationships with friends are not significant for the SWB of German children, which could point to a specific pattern of relationships with friends in Germany that may be associated with the time for enjoyment they have with friends. This may indirectly indicate that in Germany, time spent with friends is not meaningful, either because it is supplanted by other types of relationship or because not much time is spent on relationships with friends. On the other hand, it has also been observed that the inverse relationship between mistreated peers (bullying) and SWB is not significant in Norway compared to other European countries, especially compared with the United Kingdom, where this association is very significant. This could indicate that in Norway, this type of bullying situation in school is less frequent, as shown in the Table 5. The Children’s Society’s annual Good Childhood report, carried out in collaboration with the University of York, draws an alarming picture of children’s experiences at school in England and their wider sense of well-being (Rees and Main 2015). According to this report, 50% of children in the UK, 33.5% in Norway, 32.2% in Spain and 24.1% in Germany report having been left out by other children in school classes at least once in the last month. These data confirm the results obtained in our analysis regarding the differences in SWB between countries in relation to bullying. These findings could point to different educational policies to neutralize bullying, with Norway, along with Finland, being a leading country in these successful practices. In fact, the first anti-bullying program was the Norwegian nationwide campaign against bullying. This was launched through mass publicity in 1983 and involved a survey in schools, resources for teachers (curricular materials and videos) and advice for parents. It was developed by Olweus into the Olweus Bulling Prevention Program (OBPP) (Olweus and Limber 2010), which has school-level, classroom-level, individual-level, and community-level components. The success of the OBPP in Norway is thus substantial and well-replicated (Thompson and Smith 2012).

Most previous research on children’s relationships and happiness has not been based on a comparative representative sample at the European level introducing the sociodemographic factors. An important contribution of this study is that by using the International Survey of Children (ISCWeB) for several European countries, our findings can be generalized to analyze the implications for child policy in a comparative and sociological perspective.

The results provided by this study have important implications for ongoing research and for families and schools. On the one hand, the results show the positive role of good relationships with family, friends, the school and the community in children’s SWB, which corroborates the theory that social capital is a key factor for SWB (Helliwell and Putnam 2004). Second, the type of family the children live in seems play a role in the involvement of the parents and therefore SWB, with well-being being greater among children who live in families with both parents and in families where both parents work; nevertheless, these relations become less evident when the analyses by country are disaggregated.

Finally, the variation in SWB when analyzing the results by country points to possible differences in the application of work-life balance policies, in educational policies and in pedagogical programs to combat bullying, with Norway standing out as an example of a country with a more egalitarian welfare system and more advanced and innovative equality programs and policies. It would be useful for future research to examine the relationship of culture and institutions with family dynamics, the educational system and children’s interactions in relation to SWB. These findings have important policy implications for policy makers and researchers interested in the well-being of children in Europe.

The study has some methodological limitations due to the validity of the measures of children’s SWB (Pollard and Lee 2003) and the limited sociological variables included, more specifically, we do not have information on the education of the parents, type of relationship, number of siblings, economic situation or quality of the relationship. Finally, cannot draw any direct causal inferences from these findings, as the study design in cross-sectional. Nevertheless, in comparative terms, the results provide a specific contribution to social research on issues of social policies targeting children. Future research should assess the relationship between family structure, parental involvement, quality of relationships and SWB in different cultural and institutional contexts.

These results suggest that there could be a large variation within countries that needs to be taken into account. If multilevel modeling had been carried out in each country, considering regional differences and the age and sex of the children, the differences within each country might have provided new evidence on the need to develop childhood policies focused on each group of children in different regions with specific problems and needs. Therefore, based on the results of this research, we suggest further research in each country, particularly among children in these transitional periods.

Along with the findings of several studies (Lee and Yoo 2015; Dinisman et al. 2017), our findings suggest that it is necessary to provide greater attention to the underlying meanings of different patterns of relations between children’s well-being, parenting, friends and school in each country. Despite the growing attention of international organizations to the importance of child policy, few studies have analyzed the association between child welfare and children’s relationships with parents, school and friends and peers, taking into account the social policies implemented in each country. In particular, Norway emerged as a country characterized by more time dedicated by parents, school life and lower occurrence of bullying, all dimensions that are positively associated with children’s well-being. The explanation for these results may be in the innovative policies that have been developed to favor the work-family balance of parents so they can spend more time with their children and in the educational policies that promote an adequate school climate for learning and the relationship between classmates. In addition, our findings suggest that more research on the broader family and school national context is necessary to understand the potential variability of these correlates of children’s well-being depending on the context of child and family policies. The findings of our study points to the need for considering the national child and family policy context to improve children’s subjective well-being. Our findings also indicate that these policies should be integrative in the studies about child well-being to produce findings that guide public policies aimed at the well-being of children and adolescents. Therefore, our study suggests that reducing barriers to work-family balance and developing educational services in schools could be very beneficial for children’s well-being.

In summary, our study provides new analytical evidence in line with the work of Klocke et al. (2013) to try to show how children’s subjective well-being is a complex phenomenon that related to multiple factors that should not be explained solely in terms of macro social factors of an economic nature, but rather from a holistic perspective that allows us to integrate cultural, economic and social factors into an integrated interpretive framework and explain variations between nations. Although the results presented in this paper are not conclusive regarding the effects of family structure, they provide support to the beneficial role in children’s subjective well-being of satisfactory relationships with parents, friends and the school, such as having time to play and living in a safe environment and a safe neighborhood. These findings could serve as a starting point for researchers to focus on existing differences in children’s subjective well-being in European countries that are similar in principle but culturally and institutionally very different, applying a holistic interpretative perspective for this purpose. The results presented here illustrate the possibility of comparing these relationships across countries to learn more about similarities and differences in children’s subjective well-being.

The study has limitations due to the explanatory weakness of the independent variables. The survey used has specific objectives focused on psychology and neglects the introduction of sociodemographic variables. This has hampered the sociological analysis proposed in this article. For example, it is impossible to know directly from available variables the labor situation of the mother and father, the educational level of the parents and the economic situation of the household, although we have indirectly approached to these dimensions. In addition to these limitations is the fact that the questions available to measure the relationships and communication between parents and children are quite imprecise. These methodological limitations limit our approach to the question of how family communication interact with well-being of the child. However, despite these methodological limitations, the results obtained have interesting contributions for child support policies. Although the correlation indices presented here have interpretative limitations, the results obtained allow us to measure in an approximate way the possible degree of relationship between the examined variables. In addition, the selection of the variables was guided by a rigorous theoretical analysis. According to Gupta (1993) and Badu-Nyarko (2011), the selection and adequate theoretical motivation guiding the variable selection helps the variables helps interpret the results of the correlation indices obtained while acknowledging their limited statistical significance.