Florien Kruse Stagnation or transformation of attitudes towards the nurturing father ? Comparative analysis in nine European countries for 1999 and 2008-2009

There is a large body of cross-national research on attitudes towards maternal employment. However, little cross-national studies exist mapping attitudes towards father’s involvement and how they evolve over time. The aim of this paper is to enhance our understanding of the factors that relate to the attitudes towards father’s involvement by providing a rich comparison over time, country and inter-country factors by using the European Value Survey that included the following statement in 1999 and 2008-2009: In general, fathers are as well suited to look after their children as mothers. Special focus will be put on the influence of the welfare regime, age and gender. A descriptive analysis over time with age is given, to identify an intra-cohort change or cohort replacement, to explain the change of attitudes. Overall, most countries do not experience a significant absolute shift, only Great Britain, France and Spain become more supportive. Furthermore, attitudes towards fathers’ involvement seem to fit the standard welfare typology. However, Spain and Poland were much more progressive than expected. The cohort replacement effect seems to be prevalent in Spain and France. Education level and having children only seem to be of influence in the attitude formation by women. Zusammenfassung: Zu den Einstellungen zur Müttererwerbstätigkeit liegt umfangreiche länderübergreifende Literatur vor. Es gibt jedoch nur wenige länderübergreifende Studien, die die Einstellungen zur väterlichen Beteiligung an der Pflege und Erziehung der Kinder abbilden und darlegen, wie sich diese im Zeitverlauf entwickelt haben. Dieser Artikel verfolgt das Ziel, unser Verständnis der Faktoren, die mit den Einstellungen zur väterlichen Beteiligung zusammenhängen, mithilfe eines umfassenden Vergleiches von länderspezifischen und länderübergreifenden Faktoren über die Zeit zu erfassen, indem der European Values Survey genutzt wird, in welchem sich folgendes Statement in den Befragungen der Jahre 1999, 2008 und 2009 findet: „In general, fathers are as well suited to look after their children as mothers“. Ein besonderes Augenmerk wird dabei auf den Einfluss der Wohlfahrtsstaatsregime, das Alter und das Geschlecht der Befragten gelegt. Eine altersbezogene deskriptive Analyse über den Zeitverlauf wurde durchgeführt, um einen Einstellungswandel innerhalb von Kohorten oder bei einem Kohortenaustausch zu identifizieren. Insgesamt kann es in den meisten Ländern nicht zu signifikanten absoluten Verschiebungen, nur in Großbritannien, Frankreich und Italien wurde mehr Unterstützung für fürsorgliche Väter gezeigt. Darüber hinaus scheinen die Einstellungen gegenüber die Väterbeteiligung der Standardtypologie der Wohlfahrtsstaaten zu entsprechen. In Spanien und Polen waren sie jedoch fortschrittZeitschrift für Familienforschung, 29. Jahrg., Heft 2/2017, S. 228-248 229


Introduction
For years, the family-work balance debate was considered a female topic. Researchers and politicians were mainly concerned on how to foster and support an increase of women in the labor market and to which extent this participation was hampered by the welfare state. Only recently, fathers have come into the focus of both research and social policy since the limitations of a state and/or market provision of care have become visible. Over the years, scholars gained a better understanding of the concept of fatherhood and how father's involvement relates to the wellbeing of the child. However, we still have a narrow understanding of the phenomena of contemporary fatherhood. A well-explored research field is the gender gap of time spent on housework and childrearing (e.g. Shelton 1990;Bianchi et al. 2000;Sayer 2005). Yet, this is only one piece of the puzzle, as multiple other factors influence or account for father's involvement. Another important piece are the attitudes towards fatherhood. There seems to be void in research regarding the attitudes towards father's involvement, although several scholars have highlighted the necessity to gain a better understanding on how values and attitudes towards fatherhood are shaped and changed over time (Lamb 2008;Tanfer/Mott 1998;Marsiglio 1995). "Future attempts to advance fatherhood scholarship should be initiated on. [..] fatherhood stereotypes, ideals and symbolic representations" (Marsiglio 1995: 12).
The objective of this paper is to identify potential determinants of attitudes towards the nurturing father, by providing a rich comparison across time, age and country. The attitudes are put in a context of the various welfare states in order to observe if their citizens differ from one another in their support. Then it shall be pinpointed if their countries transform into supporting the nurturing father or stagnate in their attitudes. In order to identify certain drivers for changes in attitudes, the positions of potential groups, based upon age, gender and education, are mapped over time.

Historical development in defining and operationalizing fatherhood
The concept of fatherhood falls under the umbrella of 'gender ideology', which means "the individual levels of support for a division of paid work and family responsibilities that is based on this notion of separate spheres" (Davis & Greenstein 2009: 89). The main idea behind the concept of gender ideology is that gender is socially constructed by daily interaction, also referred to as "doing gender" (West/Zimmerman 1987). Gender is reproduced out of existing normative constructions. These constructions are in place to organize the social system. Social arrangements -like the family -produce a connection between gender roles and what was formerly associated with biology, like the role of mothers.
The conceptualization and understanding of fathers' involvement is a dynamic process and has been revisited by various scholars over time. One of the perspectives on fatherhood is the dichotomy between the involved nurturing father versus the uninvolved father who ignores his paternal obligations. Furstenberg (1988) argues that the dichotomy good dad-bad dad complex -emerged in the public discourse -that in itself produces a divergence between the two groups of fathers. One group of fathers is involved more closely with their children, while the other group is more derelict. Furthermore, Pleck (1984) developed a description and classification of the historical development of the concepts and ideas around fatherhood, particularly in the United States. This description builds on the notion that there are four dominant ideologies over time (Pleck 1984). First, the 'moral teacher or guide' was the prevalent norm; second, from the early 19 th to the mid-20 th centuries, the father became more distant as 'the breadwinner'; third, around the 1960s the father acted as a 'sex-role model'; lastly the father emerged as a nurturer around 1966, also referred to as the 'the new nurturant father'.
But culture and conduct does not necessarily have to align, there seems to be a mismatch between the changing attitudes towards the 'new nurturant father' and the actual conduct of fatherhood (LaRossa 1988).
The fact that the culture of fatherhood has changed more rapidly than the conduct of fatherhood would seem to represent an exception to the rule. However, it may not be an exception at all. What may be happening is that the culture is following conduct, but not in the way we normally think it does. (LaRossa 1988: 452) Due to this misconception of the involved father, the fathers are conceived as being more involved than they actually are.
Most researchers build upon the notion that the role of the father is in flux. However, this has not been left uncontested, for instance, Coltrane and Parke (1998) argue that the idea of fatherhood is more complex than this linear way of defining fatherhood over the years, "we suggest that claims about the uniqueness of recent developments in fatherhood are overstated and that both positive and negative family rhetoric lacks historical perspective" (Coltrane/Parke 1998: 7). According to them, fatherhood is oversimplified and has different meanings across the life course, class, ethnicity, and regions. Moreover, the changing family structures and the cultural diversity of the conception of fatherhood ask for new theoretical models of parenting (Cabrera et al. 2000).

Gender role attitudes
It is crucial to initiate by identifying how the attitudes towards fatherhood are shaped and how they might be connected to their behavior, in light of the theory of planned behaviour by Ajzen (1991). Most researchers scrutinize the actual division in the household but do not focus on the attitudes towards the role of the father specifically, which might be due to the fact that fatherhood is still an evolving discipline. Also, the positioning towards fatherhood is usually put together with other variables as indicator for gender role attitudes, but rarely discussed in itself. The identified determinants of gender attitudes can serve as guidance in order to find crucial elements that relate to this attitude. In general, women tend to be more egalitarian than men regarding gender role attitudes (Brewster/Padavic 2000;Scott 2008;Egmond et al. 2010). The lower educated, the married, older cohorts and the non-employed individuals tend to have more conservative gender attitudes (Brewster/Padavic 2000, Scott 1996, Egmond et al. 2010. Furthermore, the social acceptance of gender equality varies greatly by country, due to multiple institutional and economic factors (Inglehart/Norris 2003). In a cross-country study on the attitudes towards maternal employment, significant differences between the countries were shown, although the existing welfare typologies or policy difference do not completely correspond to the differences in the attitudes (Scott 2008). It was found that cohort, education and the participation of the women in the labor force have a significant effect on attitudes towards maternal employment. Also, this paper found that Sweden was most egalitarian, Spain and Germany exhibit similar patterns. The Dutch attitudes were found to be mixed.

Fatherhood in flux
In line with LaRossa's arguments, starting from the 1960s to mid-1990s gender attitudes became more liberal (Brewster/Padavic 2000;Egmond et al. 2010). This change in attitudes can be partially explained by cohort replacement effects (Brewster/Padavic 2000;Brooks/Bolzendahl 2004;Egmond et al 2010), meaning that the historical development of changing attitudes is mostly due to the fact that the older generations are replaced by the younger ones who hold different attitudes since they grow up in a different cultural and social context. Thus, this theory assumes that the attitudes are formed during childhood but thereafter remain consistent over the life course. In contrast, the assumption of an intra-cohort change in attitudes states that individuals are able to change their attitudes over the life course, due to changing locations or experiences (Egmond et al 2010). Overall, after the mid-1990s, the attitude change in gender roles is perceived to be relatively slow and inconsistent (Scott 2008).

Country justification
For this paper, nine European countries 1 of the twenty-eight European member states will be analyzed. The countries are chosen according to compelling institutional and cultural differences. Since there is no specific theoretical framework that classifies the position of fathers by welfare institutions, this paper utilizes research on -inter alia -the variations in women's employment patterns under different welfare state contexts.

Welfare state typologies
According to Esping-Andersen (1990), there are "three worlds of welfare capitalism" in the post-industrial political economies that have a significant impact on employment structures. The liberal type is where individual responsibilities are central and the market dominates. Public policies are there to reassure equality and a legal framework. The conservativecorporatist type is where the dominant idea is paternalist. The market is regulated, but the traditional family is central. The social-democratic type is critical toward the market and individual agency. Economic risks are shared collectively with universalistic policies.
The welfare state typology does not specifically reflect on how welfare regimes affect fathers' involvement in paid or unpaid work. Since the conservative-corporatist welfare regime assumes a traditional division of caring tasks between men and women and hampers maternal employment, it is assumed that attitudes towards the male-breadwinner ideal of the father will dominate here. The social-democratic type nonetheless promotes egalitarian ideals and therefore expects the nurturant father attitude to prevail.
Esping- Andersen (1990; developed one of the most well-known and applied welfare regime typologies, nonetheless this typology also triggered criticism by various scholars (summarized in Arts/Gelissen 2002). Of importance here are two suggested extensions of the welfare regime typologies.
First, the argumentation made by Fenger (2007) contests Esping-Andersen's view that the Central and Eastern European countries will converge into one of the three main welfare regimes. In this paper, an extension of the welfare regime is offered. Of primary interest for this analysis is the inclusion of the Post-Communist European type, since Poland and Hungary fall into this category. This type contains elements that resemble the conservative-corporatist type in their government expenditures. However, this type was able to facilitate female employment by providing childcare, child benefits and extended maternity leave. In the case of Hungary, the notion of the gendered separation of paid and unpaid work was very clearly supported in policies, especially excluding men from parental leave policies (Fodor et al. 2002). Through this description, the Post-Communist European type is likely to have a dominant positive attitude towards the father as male breadwinner, but, at the same time, is also likely to be more egalitarian than the conservative-corporatist type.
The second suggested extension of the welfare regime typology is the inclusion of the Southern European states (Ferrera 1996) to explain the behavior of Spain. Of particular use is the reflection of Trifiletti (1999) by shedding light on the gender dimension. The author argues: "If the state treats women (and other family members) principally on the basis of family roles […], in a sense it exerts control over their paid and unpaid work because of the lack of minimum provisions and because care work is taken for granted" (Trifiletti 1999: 54). Furthermore, Mediterranean welfare regimes do not protect the individual from social risks but assume that the (extended) family will cover most cases. This is in line with Scott's findings (1999) that the non-egalitarian gender-role contract is embedded in Spain due to the ideology of the traditional family life. This description leads to the conclusion that the Mediterranean Welfare regime is more supportive towards a gendered separation of paid and unpaid work, therefore preferring the male-breadwinner ideal of the father.
Instead of proposing for the inclusion of additional ideal welfare types, alternative typologies are suggested. Notable is the response of Lewis (1992;2001), who argues that the welfare regime typology of Esping-Andersen (1990) neglects the gendered understanding of the division between paid and unpaid work. Lewis derives a theoretical framework by classifying the institutional context based upon the dominance of the male breadwinner earner model. With Great Britain as main exemplar for the strong malebreadwinner states, France classified as a modified male-breadwinner country, and with Sweden supporting a weak male-breadwinner model (Lewis 1992). However, more recent research suggests a shift towards a more adult-worker model family for the United Kingdom and the Netherlands (Lewis 2001).
Equally important to mention for this paper is the criticism that the original welfare regime typology neglects the complex relationship between culture and welfare state policies that might be able to explain cross-national differences (Pfau-Effinger 2005). According to this approach: "welfare state policies are embedded in the societal context of the welfare culture (the relevant values and ideas in a given society surrounding the welfare state), the institutional system which comprises institutions of the welfare state and other central institutions, social structures and social actors, and their interrelations" (Pfau-Effinger 2005: 5). Kremer thereupon (2007) argues in the same line and suggests that by understanding the welfare regimes one must look at the ideals of care, thus how parents perceive appropriate care. The Netherlands is classified into two ideals of caring. First, it sets its policy norm for a parental sharing of caregiving referred to as the 'combination scenario' 2 , which is theoretically degendering caregiving. Yet, it also promoted the norm of the surrogate mother by subsidizing (primarily) mothers to take care of other one's children ('gastouderopvang'). The United Kingdom is showing a discrepancy, whereby the government is promoting the surrogate mother care ideal although its citizens did not support this.

Hypotheses
The hypotheses to be tested in this document are derived from the theories and institutional characteristics of the countries. The hypotheses are reasoned along the lines of the welfare research predictions that the same welfare regime types are likely to behave in the anticipated and same manner. It is important to mention beforehand how we assume that the welfare regimes being more supportive towards the male breadwinner ideal of the fa-ther, are therefore less supportive to the nurturant father. In Table 1, the countries are put in order on how they are likely to position themselves along the spectrum of being supportive towards the nurturant father and being less supportive, Sweden and Denmark being the most supportive and Spain being least supportive. The summary of their classification and comments are provided in Table 1 to justify the order of the countries.
Praised for their family friendly policies (Kreyenfeld et al. 2009).

France (FR)
Classified as conservativecorporatist welfare regime: By Lewis (1992) classified as a modified male-breadwinner country.
36.5 months, whereof 0.5 months are paid at 66%+ of earnings, including a ceiling on earningsrelated payment.
6 months. Whereof 0 months are paid. However part of a tax reduction.
Highest share of one-and-ahalf-job-per-household in comparison with the EU (Visser 2002).
4 months and 1 month is paid at 66%+ of earnings.
Family policies does not provide incentives for fathers to contribute to the caring labor (Plomien 2009 Institutional context reinforced by family solidarity, gender discriminatory (Trifiletti 1999: 15).
According to Pleck (1984) and LaRossa (1988), there is a trend towards supporting the nurturing father. However, it is also likely that this trend has stagnated, as this has been the case for the overall gender-role attitudes (Scott 2008). Both predictions are taken into account but it should be expected that there is an incremental change towards the nurturing father for most countries. However, the Dutch and the British citizens are likely to move away from the male breadwinner ideal more rapidly (Lewis 2001). The other variables included in the analysis are sex, age, education, employment status, marital status and having (a) child(ren). Regarding these, it is expected that women are likely to be more supportive towards the nurturant father since they hold more genderegalitarian attitudes. Furthermore, it is expected that older people are more conservative than younger ones. For education, the higher educated are likely to be more gender egalitarian and therefore supportive towards the nurturant father. Being employed is positively related to a progressive attitude towards a nurturant father, while housewives hold a more conservative attitude. Lastly, being married and/or having children relates to being less supportive toward the nurturing father than the ones who are not married and/or the ones who do not have children.

Dataset description
The most suitable longitudinal data to unravel the change in attitudes towards the role of the father come from the European Value Survey. 4 The European Value Survey is a crossnational and longitudinal survey program that has the objective to measure the attitudes, beliefs and ideas of citizens in Europe. For this analysis, the following statement of interest was raised in 1999 and 2008/2009 5 : "In general, fathers are as well suited to look after their children as mothers". This survey question was chosen due to fact this item was included in two different years, which makes cross-sectional and cross-national comparison over time possible. Furthermore, this question is likely to measure the attitude towards the nurturing capacities of the father, also called the nurturant father. Note that, in general, social survey questions measuring the attitudes towards fathers are scarce; this is particularly the case for longitudinal data.

Operationalization variables
The dependent variable is a binary variable where respondents could either agree or disagree with the statement that fathers are as well suited to look after the children as mothers. Thus, the attitude object is the capacities of the father in nurturing. The central independent variables are the nine countries justified earlier. The other determinants included in the analysis are: sex, age, education 6 , employment status 7 , having children 8 and the interviewees' marital status 9 . Furthermore, prior research tells us that the ones that are married are more conservative in their sex-role attitudes than the ones that are single or unmarried.

Data selection
All respondent that had missing values, who gave no answer or responded 'don't know' for the dependent variable are deleted from the sample. The other missing values for the independent variables were put into categories for missing values. In general, the 'missing values' categories do not show a significant relationship with the dependent variables in the models, only the missing values for education are significant in the logistic regression model. Nonetheless, the numbers of missing values are relatively low and therefore the coefficient should not be interpreted. It is important to note that most of the countries have missing values below 10% of the total sample, except for Great Britain having 16% of the data missing in 1999 and 11% in 2008. Sweden has also a striking 22% of the data missing in 2008, mostly due to the missing values in marital status, while in 1999 the missing values only account for 1%. See Appendix (Table 5, Table 6).

Statistical methods
Firstly, it was checked, if the ordered logistic regression is preferred over the logistic regression since the response options were ordered in: agree strongly, agree, disagree or disagree strongly. In order to check if the ordered logistic regression does not violate the proportional odds assumption, the Brant's test was applied (Brant 1990). 10 For both years, the overall chi-square value was significant which means that the assumptions of the ordered logistic regression were not met ( Table 7, Table 8). Therefore, the binary logistic regression is favored over the ordered logistic regression. Secondly, for the descriptive statistics the weights are included, but for the analysis this is not taken into account because the weights are constructed based upon gender and the age categories (GESIS 2014), and these are already included as control variables in the models.

Descriptive statistics
Before discussing the results of the multivariate analysis, the descriptive results give an indication of how the attitudes are distributed in absolute terms without controlling for specific other factors. Figure 1 shows that Sweden has the highest percentage of respondents who agree that fathers are as well suited to look after the children as mothers (91.9% in 1999 and 92.8% in 2008) in comparison with the other countries. There are three things to take away from Figure 1. First, the hypothesized ranking of countries roughly corresponds to the descriptive results. Only Poland and Spain seem to have a higher percentage to support the nurturing father than anticipated, with 85.8% Polish citizens agreeing with the statement in 1999 and 83.1% in 2008. Second, most countries seem to stagnate in their change in attitude towards the nurturing father, which does not correspond to the idea that the ideal of the father is in flux. Third, the countries that do experience a considerable shift in their attitudes are France, Great Britain and Spain. Figure 1: Percentage of respondents who agree to the statement that the father as well suited to look after children as mothers (weights included) * Note that the y-axis starts at 50% instead of 0.
Source: European Value Survey.
To identify which social groups experience a change in attitudes, this section zooms in on the variance in attitudes by the different age groups. The descriptive statistics of those groups provides a first insight on how the change of attitudes in different life stages evolves. 11 The order of the countries is expected, based on the support for the nurturing father in the mentioned welfare theories. In Table 1, both Sweden and Denmark show a slight decrease of support towards the nurturing father in the youngest age categories. While in the oldest age categories, from 55 onwards, both countries have much lower percentages that agree to the nurturing father; nonetheless, they experience a substantial shift towards the nurturing father over the years. In general, Sweden seems to have the highest percentages for almost all the age categories per year compared to Denmark.
It stands out that people of all ages in France transform their attitudes towards supporting the nurturing father, with an especially large increase of being more supportive in 11 The sample sizes of the age groups differentiated by country and year are too small to go beyond the descriptive analysis of the age groups or cohort. the oldest age categories. For the Netherlands, in 2008 the youngest age group seem to be less supportive than the ones in 1999, and for the people ageing 35 to 54, no change in their attitudes is shown in this table.
Great Britain experiences a complete attitude shift towards the nurturing father in all age groups. Especially the 45-54 age group became more supportive than their prior one the year before. Also, the same cohort in 1999 measured in the age category of 35-44 had much lower percentages. This might indicate an intra-cohort shift in attitudes.
Table 1 also shows that Hungary became more supportive over the years in every age group. Especially the youngest age group and the 54+ seem to have shifted. The youngest age group became more conservative while the older age groups show the opposite trend. In Poland, in Figure 1 the decline of Poland was noticeable; it appears that mainly the youngest age groups moved towards a more conservative attitude and only the older age groups 54+ developed a more supportive attitude towards the nurturant father.
In 1999, the youngest age groups in western Germany are more supportive compared to the ones in eastern Germany, whereas the older age groups (45+) in eastern Germany are more supportive than the West Germans. In 2008, we observe a convergence of attitudes between East and West Germans with the exception of the oldest age group. Interestingly, the youngest age group of East Germans is more supportive towards the nurturing father in 2008, than their former cohort, as well as the West Germans in the same age group and year. The difference in percentage points between the ones belonging to the youngest age group of 15-24 of 1999 and 2008 were calculated to identify if there is potentially a cohort replacement effect in the attitudes. To identify if the new generation holds significantly different attitudes, the two tailed Fisher's exact test is used. This test provides more acquired information, than for instance the Pearson's chi-square test, with relatively small sample sizes. 12 Only the p-values are shown in Table 2. Hereby, it shows that only France and Spain seem to have a new cohort with significant different attitudes than the individuals in the age group 15-24 in 1999. The descriptive statistics illustrate how intra-country differences are connected to their predicted attitudes, due to their welfare states. Also the various age structures give more insight of a replacement effect or intra-cohort attitude changes. To enhance our understanding on how the attitudes towards fatherhood are constructed further analysis is necessary to obtain the coefficients and their significance, for both the intra-country and inter-country differences which will be done in the next subchapter.

Multivariate analysis
To identify the transformation of attitudes towards the nurturing father over those nine or ten years, an interaction between the countries and years are made. Thereafter, the predicted probabilities are plotted in a margins plot, separated for men and women. The countries are positioned on the expected spectrum on how likely it is that they support the nurturing father, from being totally supportive on the left and conservative on the right. This seems to correspond to most of the countries, as it is almost everywhere a downwards line, but Poland and Spain are much more progressive than anticipated. These margins plots show that overall the predicted probabilities of men is lower to support the nurturing father than the women for all countries, since men position themselves lower on the y-axis than women. Furthermore, for both genders a similar pattern prevails over the years per country. France shows a significant difference between 1999 and 2008 for both genders and the same applies for Great Britain and Spain. For the other countries, there does not seem to be a significant difference of attitudes over the years and especially Germany seems to show constant attitudes over those nine years.
To get a better understanding of what might influence the composition of the attitudes towards fatherhood, the rest of the inter-country variable coefficients are given in the first row of the table below. The second and the third row are separated by gender, because men and women might differ in their determinants that influence their attitude towards father's involvement. For instance, for men this question might not be hypothetical since they are fathers themselves. Women 1999Women 2008Women -2009 Predictive margins by country, year and gender with 95% CIs As the margins plot already indicated, women are more likely to agree to the statement "In general, fathers are as well suited to look after their children as mothers" than men. Furthermore, the oldest age categories (55 and older) are the least likely to agree with the idea of the nurturing father. Also, the higher educated women are significantly more supportive than the lower educated ones. Interestingly, there are no educational differences for men. Literature tells us that having children makes people more conservative. This is also shown for mothers, since they have lower odds to agree with the nurturing father than childless women. Lastly, divorced, separated or single men are more likely to support the idea of the nurturing father. For divorced or separated women, the effect is in the opposite direction. The latter is likely to be due to maternal gatekeeping.

Discussion
Overall, Sweden is the country that is most supportive towards the nurturing father compared to the other countries. The countries show to a certain extent the predicted attitudes towards the nurturing father, and positioned themselves close to the same welfare regime. However, Spain and Poland were much more progressive than anticipated.
In the margins plot (Figure 2), it shows that only France, Great Britain and Spain experience a significant transformation towards supporting the nurturing father, while the other countries seem to stagnate in their attitudes formation.
The new cohort in France and Spain are significantly more progressive than the former age 15-24 age group in 1999. The older age groups had overall the lowest percentage compared to the other age groups to support the nurturing father, and also have significantly lower odds to support the nurturing father than the youngest age group. However, the descriptive percentages show that older age groups became more progressive over time in Sweden, Denmark, the Netherlands, West Germany and Poland.
Women seem to be significantly more supportive towards the nurturing father than men and for the countries which experienced a significant change over time (France, Great Britain and Spain) both genders changed simultaneously. Furthermore, the effect of education seems to have an impact on women only and not on men, the same holds for the effect of having a child, whereby only mothers instead of fathers are signifycantly less supportive than their childless counterparts. This asks for further investigation. For instance, is education or having a child considered as a driver of attitudes towards fathers or does this only hold for women?
Nonetheless, one should take into account that the interpretation of the question might differ per country and that this might explain certain variances. For example, how the word 'suitable' is interpreted in the questionnaire might deviate by country. Furthermore, one might argue that this data is slightly outdated and that countries like Germany have changed their family policies significantly and moved away from preferring the male breadwinner ideal since 2008 (Fleckenstein 2011).
This paper tried to provide a descriptive overview and does not give an in-depth explanation of the transformation or stagnation of the attitudes. Yet, the objective of this research is to trigger further research on this topic, and hopefully more contemporary data is available by then to unravel the transformation of attitudes by country and social groups.