Ideals and norms related to fatherhood in Europe: A comparative perspective from the European Social Survey

Objective: This study explores cross-country similarities and differences in individuals' perceptions of the ideal age and pathways to become a father, focusing on five European countries: Germany, Poland, Spain, Sweden, and the UK. Background: Empirical research analyzing family-related social norms and individuals' preferences is still scarce when compared to the abundant literature on family behavior, and especially so when focusing on men rather than women. This study attempts to mitigate this gap in the literature by focusing on ideals and norms related to fatherhood. Method: Using European Social Survey data from the most recent available round (2018/2019), descriptive and multivariate regression analyses are performed to examine: (a) the ideal age to become a father; and (b) approval of men’s decision to never have children, to have a child outside marriage, and to keep working full-time when having small children. Results: Findings confirm signs of convergence across countries regarding the “normalization” of postponed fatherhood, as well as increased detachment from traditional attitudes. Differences between “forerunner” and “laggard” countr ies with regard to family-related norms and family change are visibly narrowing. However, the ideal age for fatherhood and the approval of non-traditional life course trajectories also reflect different incentives and possibilities for the establishment of new family models provided by the gender culture and the welfare regime in each country. Conclusion: Examining social norms regarding male reproductive decisions and the exercise of fatherhood from a comparative perspective is important for understanding men’s choices and the normative social framework potentially constraining them. that the ideal age for fatherhood and the level of social approval of non-traditional life course trajectories are culturally but also institutionally embedded. Observed attitudes reflect different in the but also the different incentives and possibilities that the gender culture and the welfare regime provide for the establishment of new family models in each country. Although both show the highest degree of approval of male deliberate childlessness and out-of-wedlock fatherhood at the normative level, the generous social protection of the Swedish welfare facilitates autonomous, non-normative behaviors, while greater reliance on the family as a


Introduction
The past fifty years, and particularly the first decades of the 21 st century, have been characterized by deepgoing transformations regarding family behavior all over Europe. Among the most noteworthy changes have been the diffusion and consolidation of cohabitation, the delay and decline of marriage, the rise in non-marital childbearing, the growing plurality of family configurations and trajectories, and the steady fall and postponement of fertility (Sobotka & Berghammer 2021). Shifts regarding the timing and sequencing of family-related events have also been substantial. There has been a move from what was termed "early, contracted, and simple life course patterns" to more complex trajectories entailing later events that occur over longer periods . At the same time, there has been an evolution regarding gender relations and the roles expected of individuals throughout their life course and within the family (Pailhé, Solaz & Stanfors 2021).
In this context, questions arise about individuals' attitudes towards reproductive decisions and the exercise of parenthood. The rise in women's educational level and employment participation has fostered the delay in family formation and has set in motion the erosion of traditional gender behaviors (Pailhé, Solaz & Stanfors 2021). Economic uncertainty and widening socioeconomic disparities have prompted increasingly diverse family trajectories (Perelli-Harris et al. 2010). The wide availability of modern contraception and expanded access to assisted reproductive techniques have also had an evident impact on fertility patterns (Liefbroer 1999). The noted transformations of family dynamics, encompassed within the framework of the "Second Demographic Transition" (SDT) (Lesthaeghe 1983;van de Kaa 1987), have been linked to ideational change and secularization, increased individualism (Beck & Beck-Gernsheim 2002;van de Kaa 2003;Lesthaegue 2010), and a resulting weakening of traditional family norms (Hofäcker & Chaloupková 2014), even though social expectations about family behavior still prove relevant for individuals' choices Van Bavel & Nitsche 2013;Liefbroer, Merz & Testa 2015). Notwithstading these theoretical arguments, empirical research analyzing family-related social norms and individual preferences is still relatively scant, particularly when compared to the abundant literature focused on family behavior and its socioeconomic correlates .
This paper seeks to contribute to existing knowledge on family-related norms and values in several respects. First, we place focus on men's roles in contemporary family dynamics by looking at men's and women's perceptions of fatherhood-related ideals and norms. There is a gap in the literature that calls for evidence on the topic, as earlier research on family-related decisions has paid greater attention to women's choices and views (Hviid Malling et al. 2020). We look at timing norms (perceptions on the ideal age to become a father), quantum norms (approval of voluntary male childlessness), sequencing norms (approval of out-of-wedlock fatherhood), and involvement norms (approval of men's decision to work full-time when having children under age 3). We expect this analysis to further our understanding of men's role in the reproductive sphere by providing insights on the social norms that are likely to influence their behavior.
In addition, by analyzing five European countries -which represent different welfare regimes, in which the intersection of state, market, and family shape various family formation ideals and life course expectations-, we can assess whether there are indications of convergence or divergence in general perceptions on the ideal age and pathways to fatherhood. Earlier cross-national research on the evolution of family values found differences in age norms and acceptance of non-standard models depending on countries' progression in the SDT (Liefbroer, Merz & Testa 2015). Nevertheless, such findings relied on data from the mid-2000s. Furthermore, it is plausible that patterns have since become more similar due to the SDT's progression throughout different regions, the diffusion of norms and behaviors across increasingly interconnected European societies, and the increasing levels of educational attainment, particularly among women. On the other hand, the different institutional contexts and cultural legacies may still encourage divergent or non-uniform developments in the attitudinal and normative realms. We look into these alternatives using data from the most recent wave available (2018/2019) of the European Social Survey.
Finally, we pay attention to differences between individuals with and without college education, as well as between those below and above age 30. Highly educated individuals frequently hold less traditional values than those with lower education (Grunow & Evertsson 2019). Furthermore, a prolonged educational period pushes the age of reproductive decisions upward, possibly contributing to the mainstreaming of parenthood postponement. Less educated individuals may show preferences reflecting greater exposition to economic and labor market uncertainty, as would younger individuals. The latter may also show greater approval of non-traditional behaviors having experienced diffusion over the past decade, such as nonmarital cohabitation.

Norms and values as guiding principles of demographic behavior
The importance of social norms has long received attention within the sociological and demographic literature (Durkheim 1983;Lesthaeghe 1980). Over the past decades, there has been considerable discussion regarding their centrality in explanations of demographic behavior. The theoretical framework known as the Second Demographic Transition (SDT) (Lesthaeghe 1983;van de Kaa 1987) posited a process of ideational and behavioral transformation grounded on increasing individualization, secularization, and weakeningthough not necessarily disappearance-of normative constraints (Beck & Beck-Gernsheim 2002;van de Kaa 2003;Lesthaeghe 2010).
Empirical research has found, nonetheless, that social norms and expectations continue to play important guiding roles in demographic behavior Radl 2016;Allendorf et al. 2020). They fulfill a psychological function and help individuals structure their life course by providing orientations about the timing, sequence, and context deemed adequate for different life course transitions (Van Bavel & Nitsche 2013;Liefbroer, Merz & Testa 2015). This entails that individuals in a given society share common cultural values concerning when and how family transitions are to be performed (Settersten & Hägestad 1996). Social norms can be either prescriptive or proscriptive -that is, they indicate what must or must not be done-, they are usually backed up by a generalized societal consensus, and their compliance is reinforced by different social control mechanisms (Settersten & Hägestad 1996). As regards norms specifically related to demographic behavior and family transitions, a distinction has been made between timing norms -socially shared expectations about the age at which events should occur (Allendorf et al. 2020)-, quantum norms -i.e. how many times an event should or should not be experienced )-and sequencing norms -i.e. the order in which those events should ideally take place (Settersten & Hägestad 1996).
Social norms concerning men's and women's behavior are likely to differ from each other, as children are socialized into distinct gender roles from an early age (Stockard 2006). Accordingly, boys and girls internalize different identities, values, and expectations; and this socialization continues into and throughout adulthood. This leads to different normative standards for men and women regarding family formation. Earlier research has found, for instance, that age norms and age deadlines for the transition to motherhood are stricter than those applied to fatherhood (Van Bavel & Nitsche 2013), althoughinterestingly-men seem to face greater disapproval than women when choosing not to have children (Rijken & Merz 2014).
Social norms have probably provided blueprints for the profound demographic transformations observed over the past decades. Accordingly, we would expect to find ideational correlates of the delay in family formation and the spread of non-traditional family choices observed in European societies. As regards timing norms for men's family behavior, we anticipate higher ideal ages for fatherhood in countries with higher men's actual mean age at first birth. When it comes to quantum norms, we expect to find high levels of social approval of male deliberate childlessness in societies with comparatively high rates of childlessness (Kreyenfeld & Konietzka 2017), although research in this area is still scarce. As to alterations in the traditional sequence of family formation, the spread of non-marital childbearing across Europeespecially among individuals with low and medium levels of education (Schnor & Jalovaara 2020)-suggests that we will possibly find increased tolerance towards fatherhood outside marriage.

The SDT and the weakening of traditional norms governing family life
The narrative of the SDT links the demographic transformations observed since the late 1960s in postindustrial societies to a concomitant and deep-going shift in values (Lesthaeghe 1983;van de Kaa 1987). The latter would have stemmed from modernization, secularization and individualization processes resulting from economic and social developments -e.g., women's labor market incorporation-and would have entailed a significant increase in the importance assigned by individuals to autonomy and self-fulfillment (Beck & Beck-Gernsheim 2002;van de Kaa 2003;Lesthaeghe 2010). A progressively more individualistic context would have facilitated detachment from traditional norms and the emergence of a plurality of innovative behaviors, such as the postponement of family building, the establishment of new partnership forms, the increasingly complex family relationships after union dissolution, or the conscious decision not to have children.
Nonetheless, the progression of the SDT and its impact on norms regarding sexuality, family formation and reproduction has been shown to vary considerably across societies and cohorts (Lesthaeghe & Moors 2002). The Nordic countries and the Netherlands are regarded as forerunners in the transition towards a diminished importance of traditional norms. The SDT gradually diffused over other societies in Western Europe and North America (Lesthaeghe & Neidert 2006). Ideational change started far later in Southern European countries -Greece, Portugal and Spain had to live under a dictatorial regime until the mid-70s-, but once the process of value modernization was set in motion, it was remarkably rapid and intense. Besides divergences between forerunner and laggard countries, contemporary differences can also be linked to cultural variations -e.g., derived from differences between the Protestant and Catholic traditions, or historical family formation patterns-that have facilitated or delayed ideational and behavioral demographic change (Reher 1998). There is evidence suggesting that Central and Eastern European societies also initiated the SDT before the fall of communism and transition to a market economy (Hoem et al. 2009).
As the SDT has progressively spread across European societies, we should expect to find attitudinal detachment from traditional norms about fatherhood in all countries; and hence, relatively high tolerance towards voluntary male childlessness and births outside marriage. Social acceptance of non-traditional family behaviors is likely to be most evident in countries having pioneered the SDT (Scandinavian countries) or those having soon followed (Western European countries), yet it might have attained relatively high levels as well in Southern Europe, where family change has been remarkable rapid since the 1990s (García Pereiro, Pace & Didonna 2014). Less traditional attitudes should also be most prevalent among population groups with greater exposition to values and ideologies based on autonomy and self-fulfillment. Younger individuals and those with university education are more likely to have been socialized in nontraditional norms. Consequently, we would expect them to hold more open-minded attitudes towards fatherhood outside marriage or men's decision not to have children. Conversely, religiosity tends to be associated with more traditional family-related preferences (Adserà 2006). Accordingly, we would anticipate countries where religious influence on society is still relatively strong (e.g. Poland) to have experienced slower normative change and thus to show comparatively lesser acceptance of non-traditional fatherhood and family-related behaviors.

Other mechanisms potentially encouraging normative divergence across European societies
Despite the spread of non-traditional family trajectories in Europe and their likely equivalence in the attitudinal realm, there could remain significant cross-national differences in this latter respect. As noted, divergence between countries might result from heterogeneous progress of the SDT and related normative shifts at the societal level. It has been proposed that traditional social norms should no longer exert any considerable influence on demographic dynamics in countries having pioneered the transition , while they could remain of substantial importance where the SDT is still less advanced. This view has found empirical support. Liefbroer, Merz & Testa (2015) found significant cross-country variation regarding approval of non-traditional fertility behaviors correlating with how far along the country had come in the SDT process. At the end of the 20 th century, there was indeed significant cross-country heterogeneity regarding the diversification of family trajectories. Countries with social-democratic, conservative, and liberal welfare regimes showed a clear drop in the prevalence of traditional family models, while such was not yet the case in Mediterranean and former communist societies.
At the normative level, according to studies undertaken after the turn of the century (Liefbroer & Merz 2009;Hofäcker & Chaloupková 2014), Scandinavian countries exhibited the highest degree of tolerance for departures from traditional life course trajectories, while greater resistance was found in Southern and Central and Eastern Europe. Such variations, nevertheless, might not exclusively reflect different progress in the SDT, but also -and perhaps even more substantially-the different incentives and leeway that welfare regimes provide for the establishment of diverse family models. The generous social protection of social democratic regimes would have facilitated more autonomous, non-normative behaviors, while the greater reliance on the family as a safety net, characterizing Southern Europe, would have rather slowed down the abandonment of traditional family systems (Elzinga & Liefbroer 2007). This could lead to the expectation of higher levels of approval of non-traditional fatherhood and family models in Scandinavian welfare states in comparison to Mediterranean-type ones. Liberal countries, which do not provide comprehensive social protection, might be expected to lie somewhere in between. On the one hand, the low degree of public social provision characterizing this type of welfare regime makes young adults dependent on family resources (Esping-Andersen 1990), which may limit their options. On the other hand, the relatively individualistic cultural values observed in these societies could encourage greater life course individualization (Nauck, Groepler & Yi 2017).
The current picture could nonetheless be more complex. In Central and Eastern European societies, despite cross-country heterogeneity regarding the development of post-communist welfare states and policies, social benefit levels have been generally low, and it has been difficult to guarantee adequate living standards for much of the population (Aidukaité 2010). While this could have limited diversification of individual life courses in the direction of increased self-realization and autonomy, the 1990s saw deep-going demographic transformations in the region. As the economic transition proceeded, post-communist societies experienced a large decrease in nuptiality and fertility rates, as well as a considerable rise in cohabitation and non-marital fertility (Philipov & Dorbritz 2003). This said, it must be borne in mind that some countries (e.g. Estonia, the Czech Republic, Slovenia) have been characterized by a relatively large degree of secularization and cultural liberalism, while other societies (e.g. Poland) are more conservative and still influenced by traditional and religious norms (Sobotka 2008). It seems thus more likely that we should find a lower degree of approval of non-traditional fatherhood and family-building behaviors in the latter than in the former.
Demographic developments in Central and Eastern Europe illustrate that the level of economic and social uncertainty faced by individuals also could underlie disparities across societies regarding the extent, nature, and timing of demographic and normative change. In this line, researchers pointing at the limitations of the SDT framework have noted that pronounced economic constraints since the 1980s and 1990s -as globalization, deregulation and labor market insecurity gained momentum-might, in some countries, explain increasing acceptance of non-traditional family behaviors better than ideational shifts (Perelli-Harris et al. 2010). Cohabitation has been observed to function as a union form with lesser costs of establishment and dissolution (Kreidl & Žilinčíková 2021). Hence, we might find relatively widespread acceptance of this phenomenon, as well as of parenthood outside marriage, in contexts characterized by significant labor market and economic uncertainty. The rapid increase of cohabitation and non-marital childbearing in Southern European countries such as Spain ; Domínguez-Folgueras & Castro-Martín 2013) -even more remarkable than that observed in Central and Eastern Europe (Hiekel 2014)-, could point in this direction. So could the continuous postponement of entry into parenthood, which has also been related to labor market insecurity (Kreyenfeld, Andersson & Pailhé 2012), and which might correlate with increased acceptance of voluntary childlessness and flexible timing norms for childbearing. In societies subject to a large degree of economic insecurity in recent years -e.g. Southern and Eastern European countries-, we might thus find greater adherence to non-traditional fatherhoodrelated norms than could have been expected solely on the basis of their welfare state characteristics or their cultural legacies.

Social norms and expectations related to paternal involvement in care
Beyond social norms regarding men's reproductive decisions -whether to have children, when to become a father, and in which family context-, an important dimension of fatherhood is paternal involvement in care. While it is still mothers who bear the brunt of care work, the past decades have seen steps towards greater gender equality. Fathers' participation in their children's upbringing has increased, which has been linked to changes regarding the meaning of childcare and growing awareness of the importance of investing in children. Particularly among highly educated parents, caring parenthood norms seem to be spreading and leading to greater dedication to childcare among women and men (Sullivan 2021). Individuals with higher education tend to show greater identification with intensive parenting ideals, while devoting more time to activities that develop children's cognitive and socioemotional skills (Cano, Perales & Baxter 2019). All these factors could lead to a normative evolution towards a committed fatherhood model. Higher educated men might be most willing to advocate intense paternal involvement. Findings might nonetheless vary across societies with different gender systems.
This said, a "gender equality paradox" has been observed, entailing that highly educated professionals apparently committed to gender equality sometimes fail to adopt egalitarian practices due to work-life balance constraints (Usdansky 2011). We might thus find different attitudes regarding paternal involvement depending on whether the societal context facilitates or hinders fathers' participation in care. In countries with a welfare state regime strongly committed to gender equality, such as the Nordic countries, social support for paternal involvement is expected to be stronger than in Southern Europe or in liberal welfare states, where male workers are often penalized if they use work-family reconciliation measures. Support for paternal involvement might also be lower in conservative welfare states that have until recently encouraged a traditional division of labour while rearing young children (e.g. Germany).
Material considerations and long-lived traditional standards assigning men a primary breadwinner role might also still influence attitudes towards paternal involvement in care versus work. Qualitative research on men in Scandinavian countries (Sweden, Denmark) has revealed a wish -even in these gender egalitarian societies that have pioneered the SDT-to follow the so-called "standard chronology" before having children. Completing education, entering a stable couple relationship, and attaining employment and economic stability have been put forward as important prerequisites for becoming a father (Hviid Malling et al. 2020). A relatively consolidated position in the labor market emerges in various studies as a crucial factor for the transition to fatherhood. Difficulties related to the school-to-work transition and unemployment show a strong negative impact on men's probability to have a first birth, while this is not necessarily the case for women (Winkler-Dworak & Toulemon 2007). These findings uncover the perceived importance, for men, of guaranteeing their position as economic providers. Indeed, it has been noted that many men still regard themselves fundamentally as such, despite the shift towards more egalitarian gender values, and that difficulties in fulfilling the provider role are strongly linked to the postponement of fatherhood (Shirani 2010). On this basis, we might well find that normative ideals of full-time working fathers remain relatively persistent and pervasive even in the presence of small children in most societies.
Earlier work drawing on data from the mid-2000s found considerable variation across Europe regarding norms about women's behavior with respect to work and family, while norms regarding fathers' work intensity in the presence of very young children proved weaker and less diverse (Eicher et al. 2016). We hypothesize that such will continue to be the case, although we might find that, in contexts strongly encouraging gender equality, higher educated individuals might lean towards greater balance between fathers' provider and caring roles. Interestingly, an analysis of European fathers using European Social Survey data from 2004 found that most considered that family should be their main life priority and declared that they would like to reduce their working time (Hobson & Fahlén 2009). Nevertheless, welfare state protection proved important in this respect too. Men in countries with a comparatively high degree of protection for workers and/or measures for work-family balance -Sweden, Norway, Germany, and the Netherlands-declared they would be willing to reduce their working time considerably, even if it would entail a substantial reduction in pay. In contrast, individuals in the UK -where men have often long working hours-would only reduce it to 40 hours a week, while men in Central and Eastern European countries would rather work more hours. In a similar vein, we should find greater inclination towards parttime work by fathers of small children in societies where work-family balance and employment are wellprotected, and greater adherence to fathers' role as full-time workers where men have notably long working days, work-family reconciliation policies are limited, and/or there is an insecure economic context.

Data and methodology
The analysis is based on the ninth wave of the European Social Survey (ESS 2018/19). This survey, conducted in over thirty countries since 2002 on a biannual basis, has a cross-sectional design and is focused on measuring attitudes, perceptions and beliefs regarding various social phenomena and behavioural patterns. The 2018/2019 wave includes a module (Timing of Life) specifically dedicated to the organization of the life course and the timing of major life events. In this module, a split-ballot design was used whereby one randomly selected half of the respondents were asked about their attitudes on men's life paths and the other half on women's life paths. In the analysis, we focus on women's and men's perceptions concerning men's life-history events.
The countries included in the study -Germany, Poland, Spain, Sweden, and the UK-partly represent the existing wide diversity of welfare, family and gender regimes across Europe, which substantially shape individuals' family life transitions and the meaning attached to them (Korpi 2000). Contextual factors, such as the proportion of men with temporary work contracts (highest in Poland and Spain), the level of childlessness (highest in Germany and lowest in Poland), or the share of births outside marriage (highest in Sweden and Spain) are likely to influence fatherhood expectations and timing preferences (see Annex Table  A.1). For the analysis, we use a subsample composed of women and men aged 15 to 54 years in each country: 655 (Germany), 432 (Poland), 496 (Spain), 389 (Sweden) and 572 (UK). The intersection of reproductive, marital, educational and occupational biographies within a life course perspective, which undoubtedly influences ideals about pathways to fatherhood and its timing, drives the choice of this age range.
The analysis focuses on ideals and norms related to fatherhood. First, a descriptive examination of the ideal age to become a father is conducted. Secondly, we explore the social norms related to the approval of departures from the traditional sequence of family formation. Specifically, we assess how the respondent feels about a man deciding not to have children, a man having a child while unmarried, and a man having a full-time job while he has children under 3 years of age. The analysis is performed for the entire sample, as well as distinguishing between men and women, individuals with and without university studies, and individuals younger and older than 30 years. T-tests are conducted to assess statistically relevant differences in the dependent variables by educational level, age group and gender in each country.
The descriptive analyses are complemented with multivariate analyses that seek to highlight similarities and differences across countries, controlling for other potentially relevant factors. We perform linear regression models in which the dependent variables are the age considered ideal for a man to have his first child; approval of voluntary male childlessness; approval of fatherhood outside marriage; and approval of men's full-time employment while having children under 3 years old. As noted, the latter three ratings are based on a Likert scale that measures how the respondent feels about the different statements, ranging from 1 (very bad) to 5 (very good) (see Annex Table A.2).
As controls, education at three levels (lower secondary or less, upper secondary and tertiary level), as well as age group (under and over 30 years), are incorporated into the models. We also include covariates related to employment conditions that may affect the timing and pathways of family formation: having had or not the first job before the age of 25, experience of long-term unemployment (for longer than 12 months) and having achieved employment stability (measured as having a permanent contract at the time of the survey or previously). The degree of religiosity (measured on a scale of 0 to 10) and residence in an urban or rural area are also included in the models, as indicators of more or less traditional values and exposure to more or less innovative norms. We also control for whether the respondent (male or female) has ever fathered or given birth to children and his/her legal marital status. Following the usual convention in the statistical literature (Winship & Radbill 1994), weights are applied to the descriptive analysis, but not to the multivariate analysis. The means and percentages of the examined variables are shown in Table 1. Table 2 presents the mean ideal age to become a father and approval of non-traditional norms related to fatherhood distinguishing by educational level, age group and gender. Spain shows the highest mean ideal age, closely followed by Germany and Sweden. The mean ideal age for the transition to fatherhood is visibly lower in Poland and the UK (the only countries between which we did not find statistically significant differences when performing t-tests). A plausible explanation for these differences is that the decision to have a child depends on how individuals perceive some of the risks involved and the institutional support in place. In all the countries analyzed, except Sweden, there are statistically significant differences in the ideal age by educational attainment, with the mean difference between those with and without tertiary education being large and significant in the UK and Spain at the .001 level and in Poland at the .01 level (+1.7, +1.5 and +2.8 years respectively); the mean difference is + 0.8 years in Germany at the 0.05 level. With regard to cohort or age differences, the ideal age for the onset of fatherhood reported by individuals under 30 years of age only differs statistically from that reported by individuals aged 30 and over in Sweden and the UK. Yet in Sweden, younger individuals report an ideal age for fatherhood that is one year older than that reported by individuals aged 30 and over (p < .05). Conversely, in the UK, the mean ideal age to become a father is about one year and a half lower among individuals under age 30 than among those aged 30 and over and the difference is statistically significant at the .01 level. Overall, women report a higher ideal age for men to become parents than men, but the difference between women and men is statistically significant only in Germany and Spain. The 2018/2019 data reveal widespread social approval of men's decision not to have children; the level of approval being highest in Sweden and Spain. However, it is only in Spain that we find statistically significant differences between some of the groups. Spaniards with university education display higher approval of childless men than their less educated counterparts. The difference is also statistically significant between women and men, with women showing greater acceptance. In the rest of the countries, there are no statistically significant differences by educational level, age or gender.

Descriptive results
The degree of approval of men's non-marital childbearing is again higher in Sweden and Spain. Poland lies in an intermediate position, followed by the UK and Germany. However, differences between educational and age groups regarding attitudes towards having children out of wedlock are only statistically significant in Sweden. Table 2 shows that approval of male non-marital childbearing is higher among highly educated Swedes. Challenging expectations, Swedish respondents in their 30s and older appear more tolerant of men's out-of-wedlock childbearing than their younger counterparts. On the other hand, there are no statistically significant differences between men and women regarding approval of men having children outside marriage in any of the five countries. Source: European Social Survey 2018/2019 Notes: * p < 0.05; ** p < 0.01; *** p < 0.001. Asterisks represent statistically significant differences between the two groups in each country according to two-sample t tests

Multivariate analyses
Tables 3 to 6 show the multivariate regression analyses of the age considered ideal to become a father, as well as of approval of a man not having children, having a child out of wedlock and being in full-time employment while having children under 3 years of age. Less traditional attitudes were expected to be most prevalent among individuals with university education. However, we find only partial support for this. The linear regression model shown in Table 3 confirms that, after controlling for other potentially influential factors, educational level is positively correlated with the age considered ideal for fatherhood. Both women and men with university studies report higher ideal ages for fatherhood than their less educated counterparts in all countries, particularly in the UK (2.42***) and Sweden (2.17***). The effect of upper secondary education compared to lower secondary or less on the ideal age for fatherhood is less consistent: it is positive in the UK and Germany (also in Sweden but at the .10 level) and not statistically significant in Poland and Spain. However, with the exception of the UK, where we find a negative effect of upper secondary education with a significance level of .10, educational attainment does not show any statistically significant association with approval of male childlessness in the selected countries. Neither is the approval of male nonmarital childbearing linked to having university studies, with the sole exception of Sweden.
Regarding labour-related covariates, the findings do not show a consistent pattern. Having gained access to the labour market before age 25 is associated with a younger ideal age for fatherhood (-0.75**), but greater approval of male voluntary childlessness and of men's decision to work full-time with children under age 3 in Germany. Having had a job before age 25 is also positively associated with approval of male full-time employment during the first childrearing years in the UK (0.21***). In Sweden, individuals who entered the labour market before age 25 are also more in favour of men's voluntary childlessness (0.49***) and out-of-wedlock fatherhood (0.35***). Such individuals may have a stronger work orientation and be more inclined to endorse a childless career path.
Having achieved employment stability -measured as having a permanent contract-, is positively linked to more favourable attitudes towards a less traditional fatherhood model in Spain: results show a positive association with approval of men's childlessness and fatherhood outside marriage. While this finding may seem counterintuitive -greater job uncertainty could plausibly lead to consider it easier to have children out of wedlock-, there may be a selection-related explanation, as individuals having secured a permanent contract in relatively hostile labour market contexts, like the Spanish one, are likely to be more educated and resourceful, and thus more prone to accept non-traditional fatherhood paths. Having a permanent work contract is also positively associated with approval of out-of-wedlock fatherhood in Poland and the UK.
Men and women with permanent contracts in Sweden and the UK also report higher acceptance of fulltime employment for fathers with small children. The same is true for Swedes and Poles who have experienced long-term unemployment. However, those who have experienced long-term unemployment declare a lower ideal age for fatherhood in Spain (-0.94**) and are less accepting of male childlessness in Poland (-0.32*). It could be that these individuals question certain hostile environments and practices in the labour market and seek other avenues of identification and personal fulfilment outside paid work, such as fatherhood.
As regards the relevance of age group (or cohort) for the ideal age to become a father, there is no evidence of a consistent pattern either. The multivariate analysis in Table 3 confirms that individuals under age 30 report a higher ideal age for fatherhood than their older counterparts in Sweden (1.31**), whereas the opposite holds in Germany (-1.09**) and the UK (-0.86 †). In Germany, being under 30 is also negatively associated with approval of men's decision not to have children or to have them out of wedlock Germans aged under 30 also seem to have taken a step back from older generations in their approval of men choosing not to have children (-0.22**) or having them out-of-wedlock (-0.24**) (Tables 4 and 5).
In spite of secularization trends, religiosity continues to shape normative perceptions of men's family trajectories. The ideal age to become a father shows a negative association with religiosity in Poland (-0.14**) and Sweden (-0.18**). Moreover, religiosity is associated with more traditional attitudes related to fatherhood in all countries. Tables 4 and 5 corroborate that the degree of religiosity shows the expected negative relationship with approval of male childlessness and out-of-wedlock fatherhood. In all countries, and particularly in Poland, the more religious the respondent, the lower his or her approval of nontraditional lifestyle choices for men. Poland and Germany are the two countries with the highest levels of religiosity in the subsample; they are also the countries with lowest acceptance of men's decision not to have children or to have them out-of-wedlock. Urban environments are also associated with higher approval of non-traditional norms related to fatherhood in Poland, whereas this association is, contrary to expectations, negative in Germany, Spain and the UK.
The effect of ever having given birth or fathered child(ren) is not consistent across countries and all family milestones. Already having a child is negatively associated with the ideal age to become a father in Germany (-0.94**) and the UK (-1.08**). Not surprisingly, already having a child is also negatively associated with approval of men's childlessness in all countries but Sweden. On the other hand, ever having had children correlates positively with approval of men's decision to work full-time while having small children for the Swedes and the British ( p<.10). Table 3 on the ideal age to become a father shows that marital status is particularly relevant in Spain: being single (0.94 †) or divorced/widowed (1.77**) increases the ideal age for fatherhood compared to being married/cohabiting. The same is true in Germany for unmarried respondents (0.87**) relative to married/cohabiting ones. The association is the opposite in the UK: British women and men who have not yet entered a union report a lower ideal age for fatherhood than those who are married/cohabiting (-1.16**). German divorcees and widows/widowers, along with Spaniards, are also more approving of men not having children (Table 4). The same is true for single people in Germany, but Polish singles are less accepting of male childlessness than those who are already married or cohabiting. Table 5 also shows that those who have not yet entered into union are more accepting of out-of-wedlock fatherhood in all countries except Poland. Finally, Table 6 referring to approval of full-time paid work by fathers with minor children shows a positive association with being single in the UK and being divorced in Spain. Summing up, being married/cohabiting seems associated, particularly in the more family-oriented countries, with a higher ideal age for fatherhood and a lower approval of deviations from traditional family patterns.
Regarding gender differentials, being male is positively associated with a higher ideal age to become a father, particularly in family-oriented countries with a higher presence of the male breadwinner model: Germany (1.03***), Poland (0.67*), and Spain (0.77*). Men are also more approving of male nonmarital childbearing and male childlessness than women in Germany and Spain. This is coherent with earlier research. Although men who decide not to have children generally face greater disapproval than their female counterparts (Rijken & Merz 2014), this double standard is more often endorsed by women.

Discussion
Transformations in family dynamics have been linked to the theoretical framework known as the Second Demographic Transition (Lesthaegue 1983;van de Kaa 1987) and, within it, to ideological changes, processes of secularization, increased individualism (Lesthaegue 2010) and the weakening of traditional family norms (Hofäcker & Chaloupková 2014). However, empirical research analyzing family-related social norms and individual preferences is still relatively scarce, especially when compared to the literature focusing on demographic behaviors and their relation to socio-economic variables . Moreover, most previous studies have focused more on women than on men. Drawing advantage of the Timing of Life module of the latest wave of the European Social Survey (2018/2019), we examine ideals and norms related to fatherhood in Germany, Poland, Spain, Sweden and the UK. The added value of our study is that, besides setting the focus on men, we look at selected countries representing different institutional and cultural contexts separately rather than grouped in multilevel models like in the past (Liefbroer, Merz & Testa 2015). Firstly, following expectations, our research corroborates that, at the attitudinal level, there are clear normative correlates of the delay in family formation and the spread of non-traditional family choices registered at the behavioral level in every country. As regards timing norms for men's family behavior, we find higher ideal ages for fatherhood in those countries with higher men's actual mean age at first birth (Spain and Germany) (see Annex Table A.1). The 2018/2019 ESS data reveal that Spaniards have the highest ideal age for becoming a father: 30.1 years. In 2018, the mean age of first-time fathers in Spain was 34.4 years. Of these first-time fathers, 49% had reached or exceeded the age of 35, and 19% of them were 40 or older (INE 2021). Earlier research had shown a double standard for men and women regarding the social acceptance of parenthood at older ages (Braverman 2017).
When it comes to alterations in the traditional sequence of family formation, results confirm clear tolerance towards fatherhood outside marriage in the five selected countries, possibly due to the wide spread of nonmarital cohabitation and childbearing across Europe, not only among individuals from higher social strata but also among those with lower and middle levels of education (Schnor & Jalovaara 2020). Sweden and Spain are the countries with both the highest levels of out-of-wedlock births and the highest approval of men's nonmarital childbearing. In 2018, extramarital births outnumbered births within marriages in Sweden (54.5%). Cohabitation has also become normalized not only as a phase prior to marriage, but also as a family context in which to have and raise children in Spain (Domínguez-Folgueras & Castro-Martín 2013), where about one-third of all births were to a cohabiting couple in 2018 (INE 2021). Behavioral patterns indeed find a clear correspondence at the normative level.
However, the data suggest that approval levels do not correspond linearly with actual childlessness levels, which have been increasing rapidly in most European countries (Kreyenfeld & Konietzka 2017). Germany has long been at the forefront of this development; nonetheless, results show the lowest level of approval of male voluntary childlessness among the five countries. By contrast, Sweden has a moderate level of childlessness, but the highest level of approval of men's voluntary childlessness. Eastern and Southern Europe did not have high levels of childlessness a decade ago, but are now seeing steady increases. In fact, Spain currently ranks among the European countries with highest levels of childlessness (Sobotka 2017). In this case, attitudes and behaviour are better aligned: Spain shows the second highest level of approval of men's voluntary childlessness among the five selected countries.
Secondly, and in line with the SDT theory, our findings point at an attitudinal detachment from traditional norms about fatherhood in all countries. Fertility postponement and decline have been accompanied by more permissive attitudes regarding the ideal age to become a father and relatively high tolerance towards non-traditional family behaviors. We observe that men do not feel the same pressure to have children in their 20s or 30s as previous generations, as both men and women locate the ideal age for a man to become a father at around 30. Most of them also accept men's decision not to have children or to have them outside marriage, although, as noted, in all five countries there is greater acceptance of out-ofwedlock fatherhood than of male deliberate childlessness.
As expected, social approval of male voluntary childlessness and fatherhood outside marriage appears to be more evident in Sweden -one of the countries that pioneered the SDT-than in Germany and the UK. Yet, it has attained relatively high levels as well in Spain, where the process of value modernization and family change has been remarkably rapid and intense since the 1990s. At the end of the 20th century, there were significant cross-country differences regarding the diversification of family trajectories. However, the results are in line with previous research highlighting that differences between "forerunner" and "laggard" countries with regard to family-related norms and family change have been visibly narrowing.
Our results suggest that the ideal age for fatherhood and the level of social approval of non-traditional life course trajectories are culturally but also institutionally embedded. Observed attitudes reflect different progress in the SDT, but also the different incentives and possibilities that the gender culture and the welfare regime provide for the establishment of new family models in each country. Although both Sweden and Spain show the highest degree of approval of male deliberate childlessness and out-of-wedlock fatherhood at the normative level, the generous social protection of the Swedish welfare regime facilitates more autonomous, non-normative behaviors, while the greater reliance on the family as a safety net in Spain slows down such behaviors. The liberal-type welfare state of the UK, which does not provide comprehensive social protection, lies, as expected, somewhere in between. Poland, more conservative and still influenced by traditional and religious norms, is also characterized by relatively low social benefit levels, which constrain the diversification of individual life courses. In fact, results reveal a lower degree of approval of male deliberate childlessness and out-of-wedlock fatherhood in Poland than in the former countries. The 'familistic' orientation of Germany's social policies -an example of the conservative, or corporatist, welfare cluster-, is particularly evident in the type of public spending on families, which is relatively generous but mainly directed towards income transfers rather than towards 'de-familising' childcare services and shared parenting (Esping-Andersen 1990). Among all the countries examined, the lowest levels of approval of men's decision to never have children and to have a child outside marriage are found in Germany.
Our results are therefore in line with a substantial body of research showing that the type of welfare state matters in the production, reinforcement or depowering of gender inequalities because it influences access to both concrete resources and opportunities, as well as normative definitions of what kind of care and family are 'best for the child' and what are acceptable ways to parenting (Ciccia & Sainsbury 2018). In this sense, this study has also underscored different attitudes regarding paternal involvement depending on the societal context. Interestingly, and not surprisingly, we find that normative ideals of full-time working fathers remain pervasive even in the presence of small children in most examined societies. The image of the "distant" father, traditionally associated with the patriarchal model, has been gradually giving way to a new, more active and emotionally involved fatherhood. Nevertheless, men show greater resistance to making personal or employment-related adjustments after parenthood (Bass 2014). They often fear being punished or even dismissed if they deviate from the stereotype of the "perpetually available employee" (Escot et al. 2012). Recent studies show that some fathers even increase their hours and involvement at work to better cope with the financial responsibility of fatherhood (González et al. 2018).
In a similar vein to previous research (Liefbroer, Merz & Testa 2015;Eicher et al. 2016), we find that there are relatively small differences across countries regarding the level of approval of men's full-time work when they have small children, although individual perceptions appear to be context-embedded. Poland, a country with low institutional and cultural support for the dual earner-dual carer model, shows the greatest individual approval for men's full-time dedication to paid work. Poland promotes paternal care neither institutionally through father-friendly policies nor culturally through support of new gender roles. Nevertheless, it could be that attitudes reflect economic constraints rather than actual preferences. The same might be true for the UK.
Even in the Swedish gender egalitarian society, which has pioneered the SDT, or in Spain, which has undergone very rapid and intense cultural change, levels of approval of fathers' full-time work when having small children are relatively high. It may be that men's aspirations to become more involved fathers coexist alongside the pressures for greater work attachment or, in the case of Spain, that job precariousness translate into perceptions of risks and insecurity. In contrast, Germany -comparatively the most traditional of the five countries examined concerning the approval of non-marital childbearing and men's childlessness-shows the lowest approval of men working full-time while having small children. In brief, our research uncovers the persistence of norms assigning men the role of primarily economic providers. While younger women and men have internalized more egalitarian gender values and less traditional life goals than older generations, the normative approval of men's full-time involvement in the labour market even with minor dependent children seems to remain relatively conservative.
Although innovative because it addresses a hitherto unexplored question, we need to acknowledge some limitations in our study. First, the ideal age to become a father and norms related to fatherhood are based not only on men's but also on women's responses. In order to examine men's perceptions of their own life trajectories, it would have been necessary to select only (male) respondents who respond about individuals of the same sex (men), which would have further limited the size of the samples analysed. The choice was therefore made to collect men's and women's responses about men, but this does not allow for an in-depth exploration of the norms internalised by men themselves as a guide to their life course. The second limitation is the relatively small size of our analytical samples in each country, which may have precluded the detection of statistically significant effects; hence, our results should be seen as exploratory. Third, normative survey items often face the issue of social desirability. This would be problematic if the possible over-acceptance of one or more of the selected issues varied by country, but we have no prior empirical evidence to indicate that this might be the case.
Even with these limitations, our study suggests that observed trends in the timing and pathways of men's reproductive behaviour in the selected countries have occurred in parallel with a relaxation of the social norms that structure the life course. This flexibilization is reflected in individuals' perception of greater freedom to adapt their biography to their personal and environmental circumstances instead of following traditional patterns. In this sense, this study has attempted to mitigate the gap in the literature regarding fatherhood-related ideals, in order to learn more about men in contemporary family dynamics from a comparative perspective. Examining men's attitudes is important for understanding their choices and the normative framework that potentially constrains them, as norms differ across gender and countries. Indeed, our research points to some signs of convergence with respect to the "normalization" of fatherhood postponement, but also shows non-uniform patterns in perceptions of the age, pathways and models considered appropriate for childbearing, with divergent acceptance of non-standard behaviors depending on different cultural legacies and welfare regimes. A relevant question for future research is the relationship between the obstacles and opportunities encountered during the life course itself and the adaptation of fatherhood preferences.