Religion and Family Life: An Overview of Current Research and Suggestions for Future Research

The primary aim of this paper is to offer an overview of current research into the relationship between religion and family life and to offer suggestions for future research. In order to do this, the paper distinguishes between research in which religion is a dependent variable, labelled as socialization research, and research in which religion is an independent variable, referred to as research into family functioning. With regard to these areas of research, the paper, first, offers an overview of the most important insights existing research has generated thus far, after which some important shortcomings are identified and suggestions for future research are proposed. Subsequently, the paper offers some reflections on methodology and points out some strengths and weaknesses of existing research and offers methodological advice for future research. The paper closes by discussing how research into the relationship between religion and family life may contribute to an overall understanding of religion as such.


Introduction
Why the scientific study of religion has a lot to gain from paying attention to the family is not difficult to explain.The family is the place where the intergenerational transmission of religious beliefs and practices takes place and thus is of crucial importance for the persistence and continuation of religious traditions and communities.Three observations underscore the importance of the family in this respect.First of all, in the Netherlands and other countries in Western-Europe inter-denominational switching and conversion later in life is rare [1].Most people stick to the denomination they were brought up in or they abandon faith altogether.This indicates that religious preferences, beliefs and affiliation are determined early in life as a result of parental influences.Secondly, although parents are not the only factor in this respect, they are by far the most influential.Religious transmission may also occur in religious schools or in the religious community, but compared to the influence of parents, the influence of the religious school or the religious community is weak [2,3].Finally, religion is but one aspect of the value complexes parents may transmit to their children.Parents may for instance also transmit life style characteristics, sex roles, political party affiliation and general political orientation.However, parental influence is most noticeable in the field of religion [4].These observations clearly show that parents are crucial for the intergenerational transmission of religious beliefs and practices and thus for the persistence and continuation of religious traditions and communities.This makes the family a very important and interesting place for studying religion, as it is the most basic institutionalization of intergenerational relationships [5].
In this article, my aim is to present a brief overview of current research into religion in the family, with an emphasis on correlational and effect studies, and to propose some new directions for further research as well.Next to that, I also want to reflect on methodology and to point out some methodological strengths and weaknesses of existing research.The article closes with some remarks concerning the scientific relevance of conducting research into religion and family life for our understanding of religion as such.

A Brief Overview of Current Research
The relationship between religion and the family has been studied extensively with religion mostly being the dependent variable.That is to say, in most studies the aim is to explain how religious beliefs and practices are transmitted from one generation to the next.This is the perspective I already referred to above and which is adopted mainly by sociologists of religion and religious educators.One can briefly describe this perspective as an interest in processes of religious socialization.However, there is also a line of research in which religion serves as an independent variable.In this case, the main interest is to determine if and how religion is a factor in family functioning.This line of research is mostly pursued by psychologists of religion and family psychologists.Below, I will describe these different perspectives in more detail.

Religion as a Dependent Variable: Socialization Research
Given the fact that religious communities recruit most of their new members among the children of their existing membership, the religious socialization of children in the family has been given a lot of attention.Especially in the US this topic has been studied extensively.Here, processes of religious socialization caught the interest of researchers already before the Second World War, but it became a major topic in the sixties and seventies as a result of the growing number of apostates in certain denominations [6].Thus researchers went to look for determinants of religious commitment and involvement.Inspired by these US studies as well as by similar changes in the religious climate, processes of religious socialization also became of interest to Dutch researchers.One of the first empirical and trendsetting studies in this respect was published in 1983 by Andree [7] with the catchy title Religiosity Does Not Occur Naturally (Dutch: Gelovig word je niet vanzelf).A title which clearly expresses the core finding of this study, that religiosity, which entails both personal faith as well as religious attendance, is primarily the result of the intergenerational transmission of religious beliefs and practices from parents to child.This core finding is in line with findings from several US Studies (e.g., [8][9][10][11][12]) and it has also been corroborated by several studies conducted in the Netherlands and Belgium (e.g., [2,[13][14][15][16][17][18][19][20][21]).However, although in general these studies all indicate that parents are a crucial factor in the process of faith formation of their children, they differ in detail as to how parents actually determine this process.Still, three distinct factors seem most influential here.
Of course, one obvious factor is religious upbringing.If parents raise their children in a religious way, it is more likely that their children will follow in the religious footsteps of their parents.But religious upbringing is a multi-dimensional concept, which may refer to different dimensions like: religious instruction, performing religious practices (e.g., Bible reading and saying grace) at home, parental, religious attendance, religious saliency of parents, discussing religion at home et cetera 1 .Consequently, exactly how a religious upbringing is found to affect the religious commitment of children very much depends on the way this is conceptualized in a specific study.Nevertheless, if a multi-dimensional conceptualization is used, there is evidence that especially parental, religious behavior is of importance here.That is to say, the intergenerational transmission of religious beliefs and practices is more successful if children are confronted with visible behavioral models [14,15].This does not mean that, for instance, explicit religious instruction or discussing religion at home is unimportant, but the effects of parental, religious attendance or praying practices in the family are generally found to be much stronger (see respectively [7,21]).Thus especially parents who practice religion seem to produce children who also practice religion: Findings which are clearly in line with social learning theory, which states that people basically learn by observing and imitating others [23].
Next to religious upbringing, another important factor is family climate.Several US studies have shown that positive parent-child relationships facilitate the transmission of religion from parents to child (e.g., [9,12,[24][25][26][27]). Positive parent-child relationships are characterized by a combination of parental support and control.Such a combination is also called authoritative parenting or adopting an authoritative parenting style.The latter implies that parents are responsive to the child and express warmth (support), while simultaneously monitoring the child's activities and whereabouts (control).By doing this, parents establish an emotional climate in the family favorable to the transmission of religion.Because authoritative parenting promotes a sphere of mutual compliance and cooperation between parents and child, the latter will be more willing to accept parental (religious) values [28].However, results of research conducted in the Netherlands and Belgium are more ambivalent in this respect.Hutsebaut [15] and Alma [13] indeed found evidence of the positive effect of warm parent-child relationships on the child's identification with the Christian faith and Jongsma-Tieleman [16] reported a weak positive effect of, what she called a democratic parenting style on juvenile, religious attendance.But, more recently, Vergouwen [20] also found positive effects of an authoritarian 1 For instance, a recent study of Van de Koot-Dees [22] on the religious upbringing of young children (0 to 4 year olds) in Protestant families in Amsterdam offers a rich description of the many practices undertaken by parents in this respect.In these Protestant families parents raise their children in a religious way by such diverse activities as: prayer, singing, Bible reading, holding faith conversations, celebrating Sunday, baptism and church attendance.parenting style on juvenile, religious attendance, while a study of Vermeer et al. [29] again confirmed the importance of parental, religious attendance in this respect, but not of an authoritative parenting style.
Finally, a third important factor is family structure.In existing socialization research this factor has not been given much attention, but in their US National Study on Youth and Religion Smith and Denton [12] repeatedly and consistently found, that next to parental religiosity and the emotional climate in the family also parental marital status significantly affects teen religiosity.That is to say, growing up in a broken or single-parent family negatively affects the intergenerational transmission of religion, while being raised in a nuclear family by parents who are married actually increases the change that children step into the religious footsteps of their parents.Family structure thus also is in important factor to consider, but as far as I know this has scarcely been done outside the US.At least it has not been done in the Dutch studies I discuss in this article.
Although the above is only a very concise overview of existing socialization research, it still offers a fairly good picture of what is going on in this area.In this regard, we can, first of all, conclude that existing research is largely limited by a focus on the Christian faith and by an emphasis on Christian worship attendance.Next to the fact that researchers may deliberately focus on Christian families, this emphasis is also due to the fact that members of other (minority) faith traditions are mostly underrepresented in the national samples used.As a consequence, the core question of existing research is: How can parents transmit their commitment to the Christian faith to their offspring?Secondly, this process of intergenerational religious transmission is explained best in terms of the influence of parental modeling behavior, while there is mixed and limited evidence that also family climate and family structure matters in this respect.

Religion as an Independent Variable: Family Functioning
As mentioned already, in another line of research religion is treated as an independent variable affecting family functioning.Also this line of research originates from the US.In his classic study The Religious Factor, Lenski [30] was one of the first to systematically examine the influence of religion on politics, economics as well as on family life.Since then, numerous other studies into the relationship between religion and family life have been conducted, again mostly in the US, which are classified in a very helpful overview by Mahoney [31] (see also [32]).Mahoney distinguishes between three stages of family relationships, i.e., the formation, maintenance and transformation of family relationships, and, subsequently, discusses studies relevant for each specific stage as to show how religion is found to influence family bonds.Thus religion is, for instance, found to promote getting married, to procreate or to influence attitudes about spousal roles (formation of family relationships), to influence divorce rates, sexual fidelity or parenting processes (maintenance of family relationships) as well as to affect the way people deal with domestic violence (transformation of family relationships).Of course, it is impossible here to discuss all these different effects of religion on family functioning in detail.Thus I limit myself to the discussion of one important aspect of the family maintenance phase: the effect of religion on parenting processes.I limit myself to this particular aspect, because this is the only aspect which has also been given some attention in the Netherlands.
In order to discuss the relationship between religion and parenting, I distinguish between three aspects of parenting: goals, practices and style [33].Goals refer to the desired outcomes of parenting or to the values parents want to instill in their children.When it comes to the relationship between religion and parenting goals, the focus has been on conformity and autonomy as important values to be instilled in children.Again results are mixed.Whereas Lenski [30] reported a significant relationship between religious affiliation and a preference for conformity, with especially Catholics to value conformity above autonomy, Alwin's [34] replication study of Lenski's original study indicated an overall declining impact of religious affiliation on the preference for conformity and autonomy.However, in a more recent study, Ellison and Sherkat [35] also considered the effect of certain religious beliefs next to religious affiliation and found that the preference for conformity among conservative Protestants was mediated by specific orthodox beliefs.A relationship that was also established in the Netherlands by De Roos et al. [36]  2 .
With regard to the relationship between religion and parenting practices, several US studies have focused on Christian conservatism and corporal punishment.In their overview, Mahoney et al. [32] mention several studies that established a link between Christian conservatism and positive attitudes towards corporal punishment.In addition, they also mention that these attitudes are much more the result of conservative theological beliefs than of conservative Protestant affiliation.In his aforementioned study, Alwin [34] also considered the effect of religious affiliation on the frequency of physical punishment.But although he found that, along with ethnic and socioeconomic differences, religious affiliation adds unique variance to the frequency of physical punishment, he did not find statistically significant differences between denominations in this respect.More recently the subject was again raised by Bartkowski [39], who cites research into representative, nationwide US samples showing that conservative Protestants are significantly more likely to discipline and spank their children.As regards the Netherlands, to my knowledge this topic has as yet not been studied in a scientific way, although it received growing coverage in Dutch media in recent years.
Of the three aspects of parenting, most attention has been paid to the relationship between religion and parenting style.Also results are more similar in this respect.Overall, parental religiosity seems to be positively associated with an authoritative parenting style and especially with more positive parent-child relationships.For example, Pearce and Axinn [40] found that the mother-child relationship was affected positively by several indices of religiosity, like church attendance and mothers' personal rating of the importance of religion.However, mothers' religious affiliation in itself had no independent impact on the mother-child relationship.Consequently, Pearce and Axinn concluded, that religion influences family relationships if parents are actively involved in a religious community and thus are directly exposed to religious messages concerning child rearing.But mere nominal 2 One can, of course, also argue that the relationship between parenting goals and religion works the other way round.Instead of religion influencing the preference of parents for certain values to be instilled in children, parental values may also affect the intergenerational transmission of religion itself.Voas [37], in this respect, recently assumed that the intergenerational transmission of religion is perhaps impeded by a shift in parental values from conformity to autonomy.In order to test this assumption, Groen and Vermeer [38] studied the effect of parental values on the intergenerational transmission of religion for two generations of parents.They indeed found that among the older generation of parents a preference for conformity had a positive effect on their offspring's religious attendance and among the younger generation of parents a preference for autonomy had a negative effect on the religious attendance of offspring.However, for both generations of parents the effect of a preference for conformity or autonomy almost completely disappeared when parental, religious attendance was added to the regression model.membership has no influence in this respect.This conclusion is partly supported by the findings of Wilcox [41] concerning the parenting style of conservative Protestants.He, too, found that parents' orthodox beliefs are a stronger determinant for positive parent-child relationships-in his study frequent praise and hugging-than mere religious affiliation.Especially among conservative Protestants, then, religious ideology seems to have an impact on family life, resulting in both warm parent-child relationships and, as noted already, a greater than average propensity to spank and discipline children.A phenomenon Bartkowski [39] calls the conservative Protestant paradox and which was also found in the Netherlands by De Roos et al. [36].However, in the aforementioned, more recent study of Van de Koot-Dees [22] into religious child rearing practices in Dutch Protestant families this link between orthodoxy and disciplining again was not confirmed.Both the mainline and conservative Protestant parents she interviewed strongly favored an authoritative parenting style, pictured God as a good and loving father and rejected an authoritarian parenting style 3 .
The above overview is by no means exhaustive.Nevertheless, it clearly illustrates that religion apart from being dependent upon the family is also a determinant for family relationships.Religion functions in the formation and transformation of family relationships as well as in the maintenance of family relationships.Especially when it comes to parenting, various indices of parental religiosity, like religious affiliation, church attendance, religious beliefs, religious activities, religious conservatism, religious saliency et cetera, turn out to be influential in this respect.Unfortunately, this effect of religion on family functioning has scarcely been researched outside the US.For instance, as regards the Netherlands I am only aware of three studies in which the effect of religion on family functioning is more or less explicitly considered: viz. the studies of De Roos et al. [36], Van de Koot-Dees [22] and Vermeer [44].And while some of the findings reported in US studies were indeed confirmed in the study of De Roos et al., the study of Van de Koot-Dees in some respect yielded more opposite results, while the study of Vermeer did not reveal any relationship between various indices of parental religiosity and parenting goals or parenting style.Thus it still remains to be seen if much of the aforementioned relationships also hold outside the specific religious context of the US.

Perspectives for Further Research
Although the previous section was only meant to present a brief overview of current research into the relationship between religion and the family, some shortcomings of existing research as well as possible new perspectives for further research already came forward.The latter is now discussed in more detail below.In the first subsection new perspectives are mentioned in order to enhance existing socialization research, while the second subsection is again dedicated to religion as an independent variable.

Going beyond Existing Socialization Research
Above I characterized existing socialization research by its focus on the Christian faith and its emphasis on church attendance.This characterization already reveals two important ways in which socialization research could be developed further; viz.including other religious traditions and not limiting it to institutionalized religion.Both ways of extending existing socialization research will be discussed below together with the issue of bi-directionality as a third possible extension.
To begin with, the study of religion would certainly benefit if research into processes of the intergenerational transmission of faith is not limited to Christian families.As far as I know, religious socialization research in non-Christian families is rather scarce.Searching the online contents of journals like the Journal for the Scientific Study of Religion, Sociology of Religion, Review of Religious Research, the International Journal for the Psychology of Religion, Journal of Family Psychology or the Journal of Marriage and Family hardly reveals any study into processes of intergenerational religious transmission in non-Christian families.This does not mean that attention for processes of intergenerational transmission in non-Christian families is completely absent in the international research literature.Schönpflug [45], for example, studied processes of intergenerational transmission in Turkish families living in Germany and Phalet and Schönpflug [46] did the same for Turkish and Moroccan families living in Germany and the Netherlands, but their focus was on the transmission of collectivism and individualism values and not on the transmission of religious beliefs and practices.Nor did they consider the influence of religion in this respect.However, religion was considered in the study of Phalet and Güngör [47] into Turks and Moroccans living in the Dutch city of Rotterdam.They compared first and second generation immigrants and found that on average first generation Turks and Moroccans are more involved in religious practices, like attending mosque and daily prayer, than members of the second generation.But when it comes to the subjective importance attached to religion, there are no significant differences between generations.That is to say, both young and old consider Islam of importance for their personal life and they also consider it important that children are raised according to Islamic principles.Phalet and Haker [48] furthermore looked for determinants of mosque attendance among a representative sample of Turks and Moroccans living in the Netherlands.They found that the most important determinants are gender (men attend more frequently than women), being married with children, subjective religiosity, education (higher education reduces attendance), unemployment status and generation (first generation immigrants attend more frequently than their second generation children).Except for the subjective importance attached to religion, i.e., islam, being the same for older and younger generations, most of the aforementioned results are generally in line with results from research conducted among Christians.However, two very recent studies by Maliepaard and Lubbers [49] and Van de Pol and Van Tubergen [50] yielded more aberrant results.Maliepaard and Lubbers studied the effect of parental religiosity on the religiosity of children among Dutch Muslims, but when it comes to mosque attendance they, surprisingly, only found a strong effect of parental mosque attendance on the mosque attendance of Turkish-Dutch youths and no effect on the mosque-attendance of Moroccan-Dutch youths.Van de Pol and van Tubergen also studied the intergenerational transmission of religiosity among Dutch Muslims and even considered family climate, i.e., the degree of satisfaction with the relationship with one's parents, as a possible condition affecting the transmission of religion, but, to their own surprise, they only found a positive effect for men and no effect for women.
Especially these latter studies suggest that important religious transmission mechanisms found in Christian contexts, like parental modeling and family climate, need not automatically apply to other religious contexts.Comparing processes of religious transmission between various religious traditions and, if possible, in various cultural contexts thus could tell us more about the role parents play in this process as well as about the importance of family climate.Is religious transmission predominantly a matter of parental modeling and is this indeed facilitated most by a warm and responsive family climate or are these factors typical for the intergenerational transmission of faith in Western, Christian families?In order to deal with such a question, it is very important, therefore, to extend current socialization research to include different religious traditions in different cultural contexts.Something which would also contribute to our general understanding of religion as such as this would tell us more about how various religions persist over time in various cultural contexts.
A second important way of extending current research is to include aspects of alternative religiosity.Until now, socialization research has focused exclusively on institutionalized religion.Given the aforementioned interest in the growing number of apostates since the sixties and seventies, this is understandable.However, against the background of the contemporary critical discussion of the notion of secularization and the alleged resurgence and transformation of religion in the West (cf.for instance Davie [51]), this focus on institutionalized religion may now be considered too narrow.Especially in view of predicting future religious developments, it may be worthwhile to pay attention to alternative religiosity in socialization research.In their critical appraisal of secularization theory, Heelas and Woodhead [52] claim that traditional, institutionalized religion is indeed in decline in Western countries, but that this process is at the same time accompanied by a growing interest in, what they call, the holistic milieu.But where does this interest come from and, more important, will this interest persist?This question is extremely relevant for socialization research.As mentioned already in the introduction to this article, the continuation and persistence of religious traditions depends on the intergenerational transmission of religious beliefs and practices and it is very likely that this also goes for alternative religiosity.But in what way?Heelas and Woodhead present two possible scenarios.The "declining sacred capital scenario" relates to their empirical finding that the majority of people engaged in the holistic milieu, almost 75%, is over 45 years old of which almost 60% has been raised in a religious way.Thus the holistic milieu especially seems to appeal to people who were socialized in the Christian faith during their youth and who subsequently became disillusioned with the church.Consequently, if a religious upbringing at home is an important prerequisite for one's later involvement in alternative religious and spiritual practices, the holistic milieu will suffer momentum as fewer and fewer parents raise their children in a religious way [53].The alternative 'cultural transmission scenario', which Heelas and Woodhead consider more likely, claims that alternative religiosity and spirituality has by now become so much part of mainstream culture, that present-day youngsters encounter holistic ideas as entirely acceptable and plausible.So, while the older generation looked for an alternative for institutionalized religion, the younger generation grows up in a cultural environment which stimulates an interest in holistic spiritualities, Heelas and Woodhead [52] argue.
Thus in view of the future and alleged transformation of religion, it is of great interest to the study of religion to extend existing socialization research by testing these scenarios.Several interesting research questions emerge in this respect.For instance: Is it indeed the case that those who are currently involved in alternative religiosity and spirituality were raised in a religious way by their parents?Is there a difference in religious upbringing between those who drop out of church and lead secular lives and those who drop out of church and turn to the holistic milieu?Do parents who are themselves involved in the holistic milieu also socialize their children in this milieu?et cetera.By addressing questions like these, socialization research may offer a better insight into the mechanisms underlying the processes of religious and cultural transmission and change and thus may again contribute a lot to our overall understanding of religion as such.
Finally, the third possible extension I would like to discuss relates to the issue of bi-directionality.Classical socialization research is unidirectional.That is to say, it only considers the influence of parents on their children.This is still done in current research into religious transmission, which mostly inquires after the effect of parental religiosity on the religiosity of offspring.However, within contemporary socialization research it is generally acknowledged that children may also influence their parents and that value transmission occurs both down and up the generational ladder [43,54].As regards the Netherlands, Roest [55], for instance, recently found that late adolescents influence their fathers' attitudes towards hedonism and their fathers' and mothers' attitudes towards work as a duty.Effects of early and mid adolescents were not found, which suggests that as children mature socialization effects may become more reciprocal.However, Roest also found that value salience plays an important role in value transmission between adolescent and parents.Adolescents tend to be more successful in transmitting a particular value to their parents if this value is of importance to them, but has lower salience for their parents.Now, it is obvious that it may be very worthwhile to apply these insights to religious transmission and to study the possible effect of late adolescents on the religiosity of their parents.Again our understanding of religion would benefit from such an extension, because in Western countries subsequent generations increasingly tend to grow up in different cultural environments especially when it comes to the importance attached to religion.Studying bi-directionality in religious transmission could perhaps show, then, that adolescents are increasingly more successful in transmitting their secular life-style to their parents than parents are in transmitting their religion.Or, if the aforementioned second scenario of Heelas and Woodhead is correct, today adolescents may be more successful in making their parents susceptible to alternative religious beliefs, than their parents are in transmitting their involvement in traditional religion.
Of course, these are still assumptions, because to my knowledge bi-directionality in processes of religious transmission has been scarcely considered in current research 4 .Nevertheless, these assumptions already show that considering bi-directionality is important in order to come to a better understanding of present-day religious and cultural developments.Thus by also extending current 4 An exception in this regard is the Dutch study of Lanser-Van der Velde [17] into religious conversation between Christian parents and their adolescent children (16 to 17 year olds) as a way of, what she calls, reciprocal learning to believe affecting both parents and children.However, the families involved in this small-scale qualitative study were all highly religious with also the children participating in church activities.The children, for instance, all participated in religious education classes organized by the church; but this not what I mean by studying bi-directionality as an extension of current research.What I have in mind, here, is the study of the influence children, and especially late adolescents, may have on the (religious) beliefs and values of their parents in cases in which the beliefs, attitudes and values of parents and children diverge.socialization research in this way, its contribution to our general understanding of religion in the modern world is again enhanced.

Family Functioning and More
As we have seen, if religion serves as an independent variable in family research the focus is mainly on family functioning.Now, in order to offer new perspectives for further research, I will discuss one important way of extending current research into religion and family functioning, after which I will suggest to study an alternative extra-familial consequence of a religious family climate: philanthropy.
Although research into religion as a factor in family function revealed interesting findings in the US, these findings barely applied to the Netherlands and, I presume, these findings will also barely apply to most other West-European countries.These differences are probably due to differences in the degree of secularization.It is well-known that the US population is more religious than the European populace [56], which could very well explain why religion is found to be a factor in family functioning in the US and far less so in Europe.But far less so is not the same as completely absent.What most US studies show, is that involvement in a religious community is a crucial factor [40] (cf.also [31] for a more critical discussion of the importance of this factor).That is to say, for religious messages about family life and child rearing to be heard parents must be socially integrated in a religious community.However, in various West-European countries the level of involvement in a religious community varies according to denomination and religious tradition.In the Netherlands, for instance, the level of religious attendance as well as the level of participation in one's religious community is higher among orthodox Christians than among the members of the mainline churches like the Catholic Church and the Dutch Reformed Church [1,57].And this also goes for Muslims.Although secularization also affects the Muslim community in the Netherlands, Muslims are still markedly more involved in their religious community than most mainstream Christians.Furthermore, as I have already mentioned above, the bond with their religious community even becomes stronger as soon as Muslims get married and are having children [48].Given such differences, then, it may be very worthwhile to focus on specific religious minority groups in future research.Or to at least make sure that specific religious minority groups, like Muslims and orthodox Christians, are oversampled in order to make comparisons with mainstream religious groups possible.
Focusing on specific religious minority groups, instead of relying on national samples in which such groups are usually underrepresented, could prevent us from jumping to the perhaps preliminary conclusion that religion is not or no longer a determinant for family functioning.Making comparisons between mainstream and more orthodox religious groups or non-Christian religious communities, probably gives us a more nuanced picture of the way religion influences family functioning.For instance, it could tell us more about specific religious beliefs that may also matter in this regard next to being involved in a supportive religious community.Therefore, the study of De Roos et al. [36], which is the only Dutch study known to me that specifically compared the parenting practices of mainstream and orthodox mothers, certainly deserves to be replicated.Needless to say that extending existing research in this way is again beneficial for our understanding of religion as such.
Apart from extending existing research into the effects of religion on family functioning, it is also worthwhile to consider alternative consequences of being raised in a religious family.One interesting extra-familial consequence is philanthropy.The relationship between religion and philanthropy is established empirically in many international studies.Religious people tend to give more money to charity and are also more inclined to engage in social activities on behalf of others in need [58].Why especially religious people are involved in philanthropy is mainly explained in terms of network-theory.Most research shows that especially religious attendance matters in this respect.Hence, it is often argued, people who attend religious services are involved in local close-knit networks in which donating to charity or engaging in voluntary work has become a social norm difficult to ignore [59,60].However, there is an underlying difficulty to this network explanation especially if it is applied to the Netherlands or other West-European countries.For several decades many West-European countries witness an ongoing process of religious disaffiliation, but as yet this has hardly affected philanthropic behavior.In the Netherlands, for example, the proportion of the Dutch population involved in volunteering remained fairly stable for the past decades, while church attendance dropped significantly [1,61].These developments are clearly not in line with the network explanation, which makes one wonder if religion perhaps facilitates philanthropic behavior in other ways than offering social networks.Could it be the case, for instance, that growing up in a religious family is an important factor in this respect?This question is scarcely addressed and I am aware of only two studies in which growing up in a religious family was considered as a possible explanation for philanthropic behavior.Bekkers [62] studied the intergenerational transmission of volunteering and found positive effects of parental religious volunteering.That is to say, being raised by religious parents who volunteer for religious associations is a positive determinant for becoming involved in religious as well as non-religious voluntary work later in life.Similarly, Vermeer and Scheepers [63] studied the effect of religious socialization on children's non-religious volunteering later in life.They too found that growing up in a religious family, i.e., more precisely a family in which the religious upbringing of children is deemed important, has a lasting effect on children's pro-social orientations even if they lapse later in life.Moreover, they also found indications that the strictness of the religious family climate matters in this respect.Being raised in a religious way by one's parents is only a favorable condition for non-religious volunteering later in life if one's parents were not too religious.Thus the production of, what Putnam [64] called, bridging social capital is not enhanced by a too strict religious family climate.
Considering philanthropy as an alternative consequence of being raised in a religious family again is an interesting extension of research into the place and function of religion in the family.The study of religion would certainly benefit from it as this kind of research would give us more precise insight into the relationship between religion and philanthropy.Several interesting research questions emerge in this respect.For example: Is the effect of religious socialization on philanthropy not a temporary effect as religious disaffiliation will probably also limit the number of parents who raise their children in a religious way?Or does the effect of a more lenient religious upbringing indicate that general pro-social values gradually replace explicit religious values as motivators for philanthropic behavior?Or is the effect of religious socialization on philanthropy actually indirect as parental religiosity facilitates more positive parent-child relationships which in turn make children more susceptible to accept parental (religious) pro-social values?These are just examples of interesting new research questions.Still, these examples clearly show that also considering extra-familial consequences of growing up in a religious family may be very worthwhile.

Some Methodological Considerations
Having presented an overview of current research and offered some suggestions for further research as well, I will now present some reflections on methodology.These reflections concern the issues of causality and reliability, the benefits of working with parent-child pairs and using a multidimensional operationalization of the concept of religious socialization.
As has become clear from the above overview, most research into religion in the family involves testing causal models.Core questions concern the effect of religious socialization on offspring's religious commitment or the effect of parental religiosity on parenting goals, practices and styles.Usually these models are tested with the help of multivariate techniques like regression analysis or structural equation modeling on the basis of single wave, cross-sectional data.This is fine as long as there is a limited time lag between the cause and the supposed effect; for instance, if one wants to determine the effect of parental religiosity on the religiosity of youths still living at home.However, for testing long-term effects other data are more appropriate.For example, if one wants to determine the long-term effect of a religious upbringing on adult religiosity or on, as I have proposed above, philanthropic behavior later in life, longitudinal or panel data are more suitable.Or to put it more strongly, only the use of longitudinal data, based on at least two observations in time, or panel data, based on the observation of the same respondents on two points in time, allows one to really test causal models and to make causal inferences.However, researchers rarely have the opportunity to use longitudinal or panel data.For instance, of all the studies mentioned in the above overview of existing socialization research only the studies of Myers [9] and Vermeer et al. [21] respectively used longitudinal and panel data.Myers studied the effect of religious socialization with the help of data based on interviews with parents in 1980 and their adult offspring in 1992, while Vermeer et al. studied the effect of religious socialization by interviewing the same respondents as youths in 1983 and as adults in 2007.Only the use of longitudinal or panel data, together with the use of multivariate techniques of analysis in order to disentangle the effects of different factors, allows for a more robust testing of causal models in this respect.
Because longitudinal and panel data are scarce, most researchers make use of retrospective data.For instance, late adolescents or adults are interviewed at one point in time about their current religious involvement as well as, in retrospect, about the religious upbringing they enjoyed as a child.This may seem a good alternative for the use of longitudinal or panel data, but the question is just how reliable retrospective data are.Opinions differ strongly regarding this question.Myers [9] for instance is very critical about the use of retrospective data as he claims that retrospective recall is affected by current behaviors and attitudes, while Bekkers [62] in his study on the intergenerational transmission of volunteering claims that retrospective reports of adults about their parents' behaviors are highly reliable.Of course, this matter cannot be settled here.Nevertheless, this discussion tells us that retrospective data should be treated with caution in family research, despite the fact that in most instances retrospective data is all we have.
However, there is a way of increasing the reliability of retrospective data and family research in general.Family research typically involves at least two generations: parents and offspring.Consequently, the reliability of data can be enhanced by collecting data from two sources: the parent and the child.Using parent-child pairs avoids the danger of so-called same-source bias and enables the researcher to compare the perspective of the child with the perspective of the parent [9,25].In addition, using parent-child pairs is also of importance in view of one of the extensions of current socialization research I mentioned above; viz.studying bi-directionality.For instance, if one wants to study if adolescent or adult children transmit a secular life-style to their parents, or acquaint their parents with alternative religiosity, it is really necessary to question parents themselves about their religious development and to compare these findings with the reports of their children 5 .
My final consideration regards the operationalization of the concept of religious socialization.As we have seen, a large part of the research into religion in the family is religious socialization research.But what exactly is religious socialization and how to measure it?As mentioned already above, various indices can be used in this respect ranging from explicit religious instruction to parents modeling certain religious behavior [22].Now, it is important in research to use a multidimensional operationalization of religious socialization and not to focus on just one aspect or to combine various aspects into a single measure.This is important because different aspects may have different effects.In their socialization research Vermeer et al. [21], for instance, found that maternal, religious attendance is the strongest determinant for juvenile, religious attendance, but when it comes to the long-term effect of religious socialization on the child's propensity to volunteer as an adult, Vermeer and Scheepers [63], using the same data, actually found a negative effect of maternal, religious attendance and a positive effect of the importance parents attach to the religious upbringing of their children.This shows that using only one indicator or using a composite measure of religious socialization, the different aspects of religious socialization in the studies of Vermeer et al. [21] and Vermeer and Scheepers [63] neatly cluster into one factor, could easily result in the researcher overlooking certain effects and drawing the wrong inferences.It is more appropriate, therefore, to use a multidimensional operationalization of religious socialization.

Relevance: Family Research and Its Contribution to the Study of Religion
As to conclude this chapter, let me now draw up the balance and address the following question: How does studying religion and family life contribute to our overall understanding of religion as such?Now, it is clear that conducting research into religion and the family offers important insights into antecedents and consequences of religious commitment.As regards the antecedents of religious commitment, there is much evidence that this is primarily the effect of the intergenerational transmission of religious beliefs and practices, while it is also shown that religious commitment may have consequences for family functioning.But there is more.In my opinion, considering the familial antecedents and consequences of religious commitment also shows that religion itself is to a large extent a social phenomenon and not so much a so-called sui generis category or natural phenomenon, 5 Of course, the problems concerning the use of single wave, cross-sectional, retrospective and one-source data are closely related.Van de Pol and Van Tubergen [50], for instance, point out that their finding, based on single wave, retrospective data, that the quality of the relationship with one's parents influences the transmission of religion could very well be a wrong causal inference.For, it could also be the case that the respondents in retrospect describe the relationship with their parents as satisfactory, exactly because they agree with their parents' religiosity!If this is indeed the case, the perceived quality of the relationship with one's parents is an effect rather than a cause of the religiosity of offspring, while the retrospective report of the quality of this relationship could also be biased by this agreement in religiosity.
which is "(…) irreducible to other phenomena and impossible to explain by means of social or psychological factors" ([65], p. 3).Especially two observations give rise to this, admittedly, somewhat reductionist claim.
First of all, a religious socialization does not have a lasting effect on religious commitment at least not in West-European countries like for instance the Netherlands [21] or Britain [53,66].Although research shows that being raised in a religious way by one's parents is (almost) necessary for becoming affiliated with a religious community, a religious upbringing still does not prevent the majority of children to disaffiliate during late adolescence or early adulthood.Thus being raised in a religious way by religious parents is a necessary, but not a sufficient condition for adult religious commitment.My second observation concerns the absence of almost any influence of religion on family functioning in the Netherlands [44].Compared to the US, in the Netherlands the influence of religion on family life is as good as absent.Only the study of De Roos et al. [36] revealed some effect of religion on family functioning among orthodox Christians.So, what do these observations tell us about religion as such?Let me address this question by offering a possible explanation for these observations.Why is it that a religious upbringing does not have a lasting effect?And why does religion hardly influence a daily and value-laden activity like child rearing?
Following Bruce [67], I would say that plausibility is the key word here.Religion needs a social plausibility structure.Adhering to certain religious beliefs or performing religious practices is only meaningful to the individual if the plausibility of these beliefs and practices is supported by the individual's wider social context.Consequently, if a child is raised in a religious family it is perfectly normal, i.e., plausible, to the child to be religious.But as soon as the child enters adolescence and expands its social world the initial plausibility structure of the family may get lost if the child enters a more secular social context.Being religious is less normal, then, which may eventually cause the adolescent or young adult to disaffiliate.As Berger and Luckmann [68] have explained, whereas parents as significant others represent the plausibility of religion in the close environment of the home and thus facilitate the young child's internalization of religious beliefs and practices, the awareness of a generalized other, i.e., the awareness that one's beliefs and convictions are shared by a wider social community, is needed in order for the child to remain committed to religion as an adult.But if this wider social community is less and less pervaded by a belief system to which most people subscribe, being religious is no longer the default position which seriously hinders the transmission of faith as well as the willingness of people to let religion inform non-religious daily activities.Considered from these theoretical notions, it is no surprise, then, that a religious upbringing has no lasting effect in countries like for instance the Netherlands or Britain which rank among the more secular countries of Europe [1] 6 .Apparently, these latter countries no longer offer the religious plausibility structure that is necessary in this respect.Similarly, it is also no surprise that in the Netherlands religion no longer influences family life except for orthodox Christians.Due to their social strength and cohesion orthodox groups create their own socio-religious plausibility structure which makes them immune, at 6 That living in a more secular social environment hinders the intergenerational transmission of religious commitment was also found by Branas-Garza et al. [69] in their study into determinants of religious disaffiliation in European and non-European countries.Secular Europe proofs to be far less supportive for maintaining religious commitment than the non-European context.least for now, to the secularizing forces of the wider society.But for the majority of the mainstream Christians, in contrast, it is no longer common, i.e., socially plausible in Dutch society, to account for parenting goals or practices in terms of one's religious convictions.
Studying religion in the family thus may contribute a lot to our general understanding of religion.Considering the major results of the research I discussed in this chapter, there is strong evidence, I believe, that religion is mainly a social phenomenon.That is to say, that people adhere to certain religious ideas, perform religious practices or allow religion to inform their non-religious daily activities, to the extent that the core beliefs of a religious system are socially endorsed.Without the wider, social endorsement of religious beliefs and practices the intergenerational transmission of these religious beliefs and practices is seriously hindered.Consequently, current research seems to suggest that the essence of religion is the social.This is a rather bold claim, I immediately admit, but some of my suggestions for future research could be used to put this claim to the test.
My plea for oversampling specific religious minority groups, like orthodox Christians and Muslims, not only enables us to study specific processes of religious socialization or specific ways in which religion determines family functioning.It also enables us to investigate to what extent the lasting effect of the intergenerational transmission of faith or the strength of the influence of religion on family functioning is indeed dependent upon the strictness and social strength of the religious community involved.For, on the basis of the foregoing, one would expect that strictness and strong group ties reinforce the plausibility of the community's religion for its individual believers, which, in turn, would positively affect processes of religious socialization and reinforce the effect of religion on family functioning.Likewise, studying the intergenerational transmission of alternative religiosity or spirituality, as I also suggested above, is of equal interest.For, if the social is indeed the essence of religion, it is also very worthwhile to investigate if and how the intergenerational transmission of more individualized instances of religion or spirituality takes place.If this fails and the interest in alternative religiosity is only typical of a specific birth cohort of middle-aged, lapsed churchgoers, as predicted by Heelas and Woodhead's [52] aforementioned 'declining sacred capital scenario', this could be interpreted as evidence that a declining religious plausibility structure also hinders the intergenerational transmission of alternative religiosity.However, if parents today are gradually becoming more successful in this respect, because alternative religiosity and spirituality have become part of mainstream culture, as Heelas and Woodhead predict on the basis of their alternative 'cultural transmission scenario', this not only confirms that the strength of people's beliefs and convictions is dependent upon the extent to which these beliefs and convictions are socially endorsed, but this also shows that plausibility structures are not fixed and are subject to change.Identifying the social aspect as the main essence of religion does not therefore necessarily imply that one is also supporting the secularization thesis and that religious decline is all that one perceives.
In sum, considering the place and function of religion in the family, offers us important insights into several antecedents and consequences of religious commitment.In addition, it also helps us to come to a more profound understanding of religion as such, by enabling us to disentangle specific family factors and socio-cultural factors in explaining religion.Therefore, the family cannot be ignored within the field of religious studies.