Religion and Fertility in Western Europe: Trends Across Cohorts in Britain, France and the Netherlands

Throughout the past decades, most developed countries have experienced substantial transformations in fertility and family formation patterns. These changes, which include delays in first marriage and first birth, as well as a sharp decrease in fertility rates, are often associated with the process of secularization and the shift of values toward greater individualization (Goldscheider 2006; Lesthaeghe and Surkyn 1988; Norris and Inglehart 2004; Surkyn and Lesthaeghe 2004).

The standard theory of secularization contends that processes of modernization, including advancements in science, technology, education and economic development, would result in gradual erosion in the importance of religion, until it loses its significance in social and private life (Bruce 2011; Wilson 1966). This process is assumed to have a major influence on changing family behaviors, since decisions about union formation and reproductive choices are increasingly guided by personal aspirations of self-fulfillment, rather than by the moral order of religious institutions (Lesthaeghe 2010; Lesthaeghe and Surkyn 1988; van de Kaa 1993).

The declining significance of religion in society has also been a common explanation to the apparent convergence into lower fertility levels among different religious denominations in the USA and in Ireland (Goldscheider and Mosher 1991; Mosher et al. 1992; O’Grada and Walsh 1995; Westoff and Jones 1979). In particular, there has been a marked decline in fertility levels among those identified as Catholics. As a result, Catholic fertility became closer to the fertility levels of other major religious denominations, a process that was described as “the end of Catholic fertility” (Westoff and Jones 1979). The narrowing fertility gap between Catholics and other denominations, alongside the secularization paradigm, has led to a diminishing interest in the role of religion in explaining demographic behavior (McQuillan 2004; Philipov and Berghammer 2007). In particular, the relationship between religion and fertility received little attention in research on Western European countries, which are considered to be highly secularized (Inglehart and Norris 2003).

It should be noted, however, that the classic theory of secularization has been widely criticized by numerous scholars, who have claimed that the decline in religion is neither as widespread nor consistent as previously assumed (Davie 1990, 2007; Greeley 2003; Norris and Inglehart 2004). According to Greeley (2003), no common pattern of secularization could be found in Europe; while in some countries religion has declined, in others it has remained relatively unchanged or even increased. Although there is a general agreement that the influence of religious institutions on society in Europe has weakened, there is an ongoing debate on the continuing influence of religious ideas on attitudes and behavior of individuals (Voas and Doebler 2011). Furthermore, some scholars have maintained that a general decline in religion does not necessarily imply a weakening effect of religiosity on demographic behavior; rather, the increasing cleavage between secular and religious individuals may have sharpened the differences between these groups in terms of social and moral attitudes, which also concern family behaviors (Kaufmann 2010; Régnier-Loilier and Prioux 2008; Stegmueller et al. 2011).

In the past decade, there has been a renewed interest in the effect of religion on fertility, and recent studies have pointed to the persisting influence of religion on fertility patterns; for example, Frejka and Westoff (2008) found that in Europe and the USA, women who identify as Protestants or Catholics had higher fertility rates compared with women who declared having no religion. Additionally, within each denomination, the more devout—in terms of service attendance and importance of religion in daily life—had larger families. Similarly, in a comparative study of 18 European countries, Philipov and Berghammer (2007) reported a positive correlation between different measures of religiosity (e.g., affiliation, practice and self-rated religiosity), and individuals’ intended and actual fertility. The positive effect of religiosity on family size received further support in a longitudinal study from the Netherlands, showing that church attendance is a strong predictor of future childbearing (Berghammer 2012). In addition, using cross-sectional data from 1985 and 1999 in Spain, Adsera (2006a) has found that fertility differences between practicing and non-practicing Catholic women have grown, since fertility decline occurred only among the latter group.

However, most studies in this topic are based on cross-sectional analysis, which estimate the relationship between religion and fertility in a given point in time. The current study uses both retrospective and longitudinal data to explore whether in the context of a general religious decline, individual religiosity has become a less important determinant of family size or whether there is actually an increasing divergence between religiously active individuals and those with lower attachment to religious tradition. For this purpose, retrospective birth histories of women from Britain, France and the Netherlands are employed in order to follow trends in religious differences in completed fertility and the transition to first birth across cohorts. In addition, a longitudinal analysis is performed to estimate the effect of religious adherence on subsequent childbearing.

This study focuses on two main dimensions of religious adherence: affiliation with a particular denomination and religious practice, as well as the combination between them. Religious affiliation is regarded as a meaningful form of identification, which represents the cultural background into which a person was born or grew up in and may have consequences on social behavior (Day 2011; Southworth 2005). In addition, religious practice—as measured by frequency of attendance at religious services—is another key measure of religious commitment and is considered to be a more reliable indicator of religiosity compared with religious belief or private prayer, since it has a public aspect to it and it involves the investment of time and effort (Storm and Voas 2012; Voas 2009). Furthermore, it is important to pay attention to the interdependencies between religious affiliation and practice; first, the intensity of religious involvement is expected to accentuate the effects associated with religious affiliation (Lehrer 2004a), and second, different religious traditions have contrasting expectations about the frequency of religious service attendance (Voas and Doebler 2011). For example, Sunday mass attendance is considered an obligation by canon law for Catholics, while for Protestants, instructions regarding church attendance are more relaxed, although it is still strongly encouraged (Bruce 2011). Nonetheless, church attendance rates may differ across countries and over time (Greeley 2003; Hout and Greeley 1987; Inglis 2007).

As a result of immigration trends, a growing proportion of people in Western Europe are affiliated with a religion other than Christianity, with Muslims being the largest group among them (Pew Research Center 2009). As a whole, those identified as Muslims tend to be more religious and also to have higher fertility rates compared with the Christian majority (Kaufmann et al. 2012; Régnier-Loilier and Prioux 2008). It should be noted, however, that non-Christian groups form less than 10 % of the population in the countries observed in this study (Pew Research Center 2012), and sample sizes for these groups are relatively small. Therefore, the main focus of the current research is on fertility trends among Christian denominations and the continuously growing population of non-affiliated individuals.

The remainder of the article is structured as follows: Sect. 2 provides an overview of recent empirical findings and theoretical explanations for the influence of religion on reproductive behavior. Section 3 includes a description of religious indicators and sociocultural context of the countries selected for this study. Against this background, the research hypotheses about the relationship between religion and fertility and the way it changes across cohorts are formulated. Then, Sect. 4 provides details on the data sets, measures and empirical procedure. This is followed by the presentation of findings in Sect. 5. Section 6 concludes with a summary and discussion of findings and the implications for future research.

Theoretical explanations for the effect of religion on fertility behavior stress the high value that most religions ascribe to family and children, alongside other fertility-related doctrines, which emphasize women’s familial roles (Lehrer 2004a; McQuillan 2004; Norris and Inglehart 2004; Sherkat 2000). It should be noted, though, that some denominations may put greater emphasis on childbearing than others. For example, the higher fertility of Catholics in comparison with Protestants was often seen as a consequence of pro-natalist Catholic teachings that forbade the use of artificial means of contraception (McQuillan 2004). Furthermore, the Catholic Church has a strict opposition to divorce, since marriage is seen as a lifelong, unbreakable commitment between men and women and as the sacred basis for family formation (Richards 2009). While Protestant Churches also promote traditional family values and the procreation of children as the main purpose of marriage, the approach of Protestant traditions toward the use of contraception tends to be more pragmatic (Creighton 2009; McQuillan 2004). However, it has been argued that adherence to church teachings on contraceptive use among Catholics has substantially weakened over time and that this may account for the sharp decrease in Catholic fertility rates and the convergence of family size between Catholics and other denominations (Goldscheider and Mosher 1991).

According to McQuillan (2004), specific social and political settings may determine the extent to which religion influences fertility patterns; for example, religious groups are expected to have a greater effect on behavior when members feel a strong sense of attachment to the religious community, or when it is considered to be an important marker of identity. Furthermore, religious norms about family and fertility are most likely to influence behavior when religious institutions have the means to communicate these teachings to their members and to enforce compliance, through formal organizations or informal social pressure. Thus, religious service attendance is assumed to both reflect and reinforce belief and commitment to traditional religious values, through the repeated exposure to religious teachings and interaction with people who share similar values (Davie 2007; McQuillan 2004).

Another route through which religious participation may influence fertility is through increased social capital among those attending religious services on a regular basis; previous studies have found that religious congregations promote the formation of social networks, where different types of informational, emotional and practical support are exchanged between members (Chatters and Taylor 2005; Philipov et al. 2006; Putnam 2000; Waite and Lehrer 2003). This source of support may affect positively on fertility decisions by reducing uncertainty and by lowering the perceived costs of childbearing. Furthermore, religious belief and practice can contribute to coping with stressors and difficulties of daily life, including those related to family formation and expansion (Chatters and Taylor 2005; Krause et al. 2001; Pargament et al. 2000). For these reasons, religious involvement and church attendance in particular is expected to promote higher fertility.

These explanations are in line with empirical studies which show that religious service attendance is a stronger predictor of fertility intentions and behavior, compared with affiliation alone (Adsera 2006a, b; Philipov and Berghammer 2007). Moreover, Adsera (2006b) has argued that following the declining influence of religious institutions in Europe, church attendance has become a more salient determinant of fertility norms among younger generations, as those who continue to go to church represent a more selective group of people who adhere to religious doctrines. A similar process is described by Davie (2007), who contends that church attendance has increasingly become a matter of personal choice, leaving a small minority of religiously active people, although with strong attachment to religious convictions. Therefore, over time, the underlying differences in fertility behavior between practicing and non-practicing individuals may become more pronounced (Kaufmann 2010; Régnier-Loilier and Prioux 2008).

The three countries at the focus of this study are considered among those that have undergone the most notable religious decline in Europe over the past century (Greeley 2003; Norris and Inglehart 2004). This is evident mainly by the generational decline in the proportion of religiously affiliated individuals and in the number of people who attend religious services regularly, especially among the cohorts born after the Second World War (Greeley 2003; Inglehart and Norris 2003; Voas 2009).

Nevertheless, these countries greatly differ from one another in terms of their religious heritage and distribution of religious groups. The religious landscape in Britain and the Netherlands is relatively diverse and comprised of different Protestant denominations alongside a substantial Catholic population. In France, on the other hand, Roman Catholicism forms the dominant religion; in 2010 the share of baptized Catholics of the total French population was estimated at 75 % (Pew Research Center 2013). By contrast, Britain has a Protestant majority, where the Church of England is the largest religious denomination, forming about a quarter of the population (Voas and Ling 2010). Roman Catholics are the second largest group, forming around 10 %, while most other Christians belong to various reformed groups (ibid). Thus, over half of the population in Britain are self-identified Christians (NRS 2013; ONS 2012). In the Netherlands, the religious divide between Protestants and Catholics is even further pronounced. As part of the pillarization system, which prevailed from the end of the nineteenth century to the mid-twentieth century, members of each denomination, or unaffiliated people, belonged to separate social, educational and political institutions (van Poppel 1985). However, the pillarization system has dissolved since the 1960s, as secularization accelerated and the proportion of disaffiliated people has increased rapidly (Knippenberg 1998; van Rooden 2003). Today, Catholics constitute around 30 percent of the population in the Netherlands, while Protestants constitute about a fifth (Berghammer 2012). The Protestant population is rather heterogeneous and is comprised of mainline Dutch Reformed, as well as more conservative groups, including Orthodox Calvinists and Evangelical denominations (Knippenberg 1998).

While the majority of people in the countries observed here identify themselves as affiliated with a particular religion, only a minority attend religious services on a regular basis. Close to a fifth of the population in Britain and the Netherlands attend religious services at least once a month while just over a tenth of the French population do so (Davie 2002; Statistics Netherlands 2012; Voas and Ling 2010).¹ However, attendance rates vary greatly across religious denominations. For example, church attendance rates are markedly higher among British Catholics than among those affiliated with the Church of England (40 % of Roman Catholics attend at least once a month compared with 18 % of Anglicans, see: Voas and Ling 2010). In the Netherlands, attendance rates are currently higher for Protestants than for Catholics, due to the dramatic fall in participation rates among the latter; while more than 70 % of Roman Catholics were regular church goers in 1970, this figure has dropped to less than a quarter by the end of the twentieth century (Knippenberg 2005). Across the same period, regular attendance among Dutch Reformed has dropped from 50 to 40 %, and among other Protestant streams (e.g., Orthodox Calvinist) about two-thirds remained regular attenders, following a drop from nearly 90 % (ibid).² These trends may be the result of the traditional divide between liberal and fundamentalist wings within the Dutch Reformed Church; on the one hand, resignation from the church has started as early as the 1880s, while on the other hand, the fundamentalist wing may have contributed to maintaining relatively high participation rates (Knippenberg 1998, 2005).

In France, religious participation rates have been and continue to be considerably lower compared with those in other Catholic countries, such as Spain, Italy or Ireland (Davie 2002; Norris and Inglehart 2004). This may be partly attributed to the historical conflict between church and the state in this country and the relatively strong anti-church sentiments among the French people (Greeley 2003; Martin 2005). Nonetheless, recent data indicate that the low rates of church attendance among French Catholics have reached a plateau and remained stable over the past decade (Pew Research Center 2013).

Compared with those identified as Christians, religious participation among members of other religions (most of whom are Muslims) is expected to be higher, since this group includes a larger proportion of immigrants from highly religious countries (Kaufmann et al. 2012). Moreover, religion may play a more significant role in the social identity of minority groups (Fetzer and Soper 2005; Southworth 2005).

As only a minority of the population in Western Europe attend religious services, these countries consist of a large proportion of ‘nominal’ religious people, who identify as affiliated with a specific denomination, but do not attend religious services on any regular basis. This phenomenon has been described by Voas (2009) as “fuzzy fidelity,” since nominally affiliated individuals are neither religious nor completely secular. Moreover, Voas maintains that this is a transient stage toward increasing disaffiliation from religion. Some scholars have argued that religious identification is becoming less a matter of adherence to church teachings and regulations, and more a matter of belonging to a shared cultural heritage (Hervieu-Léger 1990, 2000; Inglis 2007; Pace 2007). This may especially be the case within Catholicism, as increasing numbers of self-identified Catholics do not share the views of the Catholic Church leaders on sexuality, which may also account for the fall in attendance (Hout and Greeley 1987; Pace 2007). Similarly, Day (2011: 72) has concluded that many people identify as Christians simply because they were baptized or attended church when they were younger, or since it is closely associated with national or ethnic identity (Day 2011; Voas and Bruce 2004). However, compared with practicing religious people, nominally religious individuals are less likely to consider religion as an important source of guidance in everyday lives (Day 2011; Voas 2009).

Under these circumstances, as the proportion of regular church attendance is shrinking from one generation to another, we may expect this group to show increasingly distinct patterns of fertility and family formation, as they are more likely to continue adhering to traditional religious doctrines that highlight the value of family and children, alongside the fulfillment of traditional family roles (Adsera 2006a, b; Goldscheider and Mosher 1991). Moreover, behavioral features of the second demographic transition, including the delay in first birth and the shift to below replacement fertility, are less likely to appear among those individuals who are dedicated to traditional forms of religious practice (Surkyn and Lesthaeghe 2004).

Therefore, it is first hypothesized that individuals who attend services on a regular basis (at least once a month) will have larger family size compared with those who attend services less often or not at all. The second hypothesis postulates that nominally affiliated individuals (those identifying with a particular religion but do not attend services on a regular basis) would have smaller family size compared with religiously active individuals, though larger than those who stated having no religious affiliation.

The third hypothesis contends that fertility differences between regular and non-regular service attendants will increase among younger birth cohorts. Since the first group is becoming smaller and more selective, those women with stronger adherence to religious doctrines are expected to maintain high levels of fertility, while other women would experience fertility decline.

In this context, women with lower attachment to religious tradition would have increasingly higher likelihood of remaining childless. Therefore, the fourth hypothesis states that more religious women, in terms of affiliation and practice, would be more likely to experience the transition to first birth and that this gap will increase among the younger birth cohorts. The following section provides a description of the data sources and methodology that are used to examine these propositions.

This study examined the changing relationship between religion and fertility in three countries that are considered to be highly secularized. While institutionalized forms of religion have been declining in Western Europe from one generation to the next, religion continues to be an important determinant of fertility and family formation patterns in this region. By reconstructing women’s births histories, it was possible to detect differences in fertility trends by religious affiliation and practice between countries and across cohorts. Moreover, these differences are confirmed in a longitudinal analysis of the role of religion in determining future childbearing.

In accordance with the study’s first two hypotheses, both active and nominal religious adherence is linked with higher fertility levels, where the highest levels are among practicing religious women. The third hypothesis, which implies a growing fertility divergence by religiosity among younger cohorts, is only partially supported by the results; an increasing fertility gap is found between practicing Catholics and non-practicing women in France as well as between practicing Protestants (Dutch Reformed and others) and non-practicing women in the Netherlands. On the other hand, fertility differences between practicing Catholic and other women in Britain have narrowed, although an increasing divergence is evident for the most recent cohort. The fourth hypothesis about the higher propensity of more religious women to enter motherhood and the increase in that gap among recent cohorts received some support in this study; the proportion of childless women has increased the most among non-affiliated women, while it has generally declined for the practicing religious ones. However, these differences are at least partly attributed to differential marriage patterns among religious groups.

Since the relationship between religious indicators and fertility behavior varies across countries, the argument of increasing selectivity among religiously observant women is not sufficient to understand these trends. Rather, additional country-specific circumstances should be taken into account. The persistence of fertility differences by religion in France and the Netherlands may be related to social conflicts on the basis of religious identity in these countries. In the case of France, the ongoing debate about its character as a secular country and anti-Church sentiments may account for the stronger divergence between actively religious and non-practicing individuals. In the Netherlands, as a result of the historical pillarization system, religious affiliation may still constitute an important part of social identity, which could account for the differential fertility patterns by religiosity. In Britain, on the other hand, religion appears to play a less significant role in explaining fertility differences in recent cohorts. It should be noted, though, that the apparent convergence in fertility patterns between Catholic women (practicing or nominal) and non-affiliated women in Britain may be temporary, as there is evidence for an upward trend in completed fertility among Catholic women who have just recently completed their childbearing years.

Overall, it appears that, as noted by Greeley (2003), there is no one single process of secularization in Europe. Although the general direction appears to be one of decline in indicators of religiosity, each society has its own variation of this process and shows different patterns of continuity and change in the extent to which religion influences individual behavior. Future research of the relationships between religion and fertility should pay attention to the variety of secularization processes in Europe and the way these influence the changing role of religion in shaping fertility patterns.