Keywords: Kinship networks, divorce, family relationships
During the last 40 years, divorce rates in Europe have doubled (Eurostat, 2017). Despite the extensive amount of research on how parental divorce affects the relationships and well-being of children and their divorcing parents (e.g., Amato, 2000; Ahrons, 2007; Sobolewski & Amato, 2007; Albertini & Garriga, 2011; Kalmijn, 2013), little is known on the consequences of parental divorce for extended kin (Widmer, Aeby, & Sapin, 2013). Hence, we argue that the complete kinship network should be studied. In this paper we investigate which and how complete social network analyses can be applied on kinship networks. This enables us to study relational structures in kinship networks of divorced families and kinship networks of intact families.
Based on family and divorce literature, kinship networks of divorced families may differ from the kinship networks of intact families on two aspects. Firstly, argued from family systems theory (Minuchin, 1974; Cox & Paley, 1997), it can be argued that grandparents may also be affected by the parental divorce because family relationships are interdependent. Interdependent family relationships imply that changes in one relationship, for example between the divorcing parents, also affect the relationships with and between other family members. Grandparents for example may see their grandchildren less after parental divorce or the relationship between the divorcing parents and former parents-in-law may become tensioned and reduce intergenerational relationship quality.
Secondly, reasoned from a solidarity perspective, extended family members simultaneously can become important sources of support to the nuclear family (Hank & Buber, 2009). For example, grandparents may take over household tasks from the divorcing parent and they may take care more often of their grandchildren, resulting in more intergenerational contact and potentially also in better intergenerational relationships.
In order to investigate these predictions, kinship networks of 60 divorced families and 80 intact families will be analyzed, using data from the Divorce in Flanders study. The kinship networks are defined by two relational dimensions: contact and relationship quality. Contact is regarded to be a behavioral relational variable, whereas it also is regarded to be a prerequisite for any further qualitative interpretation of the family relationship. Relationship quality (extend to which the relationship is considered to be good or bad) is a more qualitative, subjective, assessment of the family relationships and hence is more likely to differ among the family roles (i.e., a mother may value the relationship with her child differently compared to the evaluation of this child). The complete directed kinship networks consist of the mutual relational reports between approximately 7 family members: one resident child, two parents and up to four grandparents. In the paper, special attention will be given on selecting the right network model to analyze these kinship networks (e.g., Bayesian Exponential Random Graph Models for small binary networks (Caimo & Friel, 2014)).