Types of social networks and the transition to parenthood

Abstract

1. IntroductionHuman beings are social actors embedded in a network of social relationships with kin and peers. A basic assumption of social interaction theory is that important decisions in the life course, such as the decision to become a parent, are not only driven by an individual's personal characteristics but are also influenced by the traits and the behavior of the people with whom individuals interact (Bongaarts and Watkins 1996; Kohler and Buhler 2001; Bernardi 2003). Consequently, a body of literature, growing since the 1980s, acknowledges the importance of social interaction and ideational factors for childbearing intentions and generative behavior (Kohler and Buhler 2001; Bernardi and Klarner 2014). This direction of research suggests that social interaction creates, diffuses, and transforms the value of children, support for parents, and social norms that regulate both the timing and quantum of fertility.Much of the previous research into the effects of social networks on fertility has been undertaken in high-fertility contexts such as Eastern Asia and Africa (Montgomery and Casterline 1993; Kohler, Behrmann, and Watkins 2001; Rindfuss et al. 2004; Sear et al. 2009). The early focus of this literature was the implementation of state programs for birth control as well as the diffusion, acceptability, and use of modern contraception. Beginning in the early 2000s, however, the social network approach also received increasing attention as a means of explaining fertility changes in western, post-industrial societies. Examples are studies of the timing and postponement of births, the emergence of low fertility, and the diffusion of childlessness (Kohler 2001; Philipov and Kohler 2001; Kohler, Billari, and Ortega 2002).Recent qualitative research suggests that the influence of social networks on fertility operates through the following four mechanisms (Bernardi 2003; Bernardi, Keim, and Lippe 2007; Keim 2011). First, 'social pressure' is the force that leads individuals to conform to social norms, such as their parents' wish for grandchildren, in order to gain approval or to avoid conflict. Second, 'social support' is the amount of emotional, instrumental, and material help parents can activate within their social network. Third, 'emotional contagion' is defined as the spontaneous pick-up of emotional states; for example, when someone feels emotional arousal when holding a friend's baby. Fourth, through 'social learning' individuals can acquire information about the costs and benefits involved with having children.Several studies in the field of social network research strongly suggest that the effect of these four mechanisms depends on the characteristics of the parents' social network, including its structure, its composition, and the nature of its relational ties (Bernardi and Klarner 2014). For example, research has shown that the reliability of Ego's2 informer, which should be higher for strong ties, is crucial for effective social learning and actual adaption (Buhler and Kohler 2004).Following this line of thought and building on qualitative research (Keim 2011), the main goal of the present study is to identify specific types of social network by variance in their structural characteristics as well as in their potential for social pressure, social support, emotional contagion, and social learning. Although studying the separate effects of each specific mechanism carries the advantage of higher analytical precision, the typological approach proposed here is more realistic because it identifies the multidimensional social configurations in which people are embedded. Moreover, a major strength of the network approach is the identification of patterns that moderate the effectiveness of each individual mechanism of social influence (cf. Keim 2011: 216).Further, the present study gauges the predictive validity of network types with regard to the transition to parenthood. The analysis employs data from the German Family Panel Study (pairfam), which include repeated measurements of the characteristics of ego-centric social networks. …