Parental background and union dissolution from a cross-national comparative perspective

Anne M. D. Brons, Netherlands Interdisciplinary Demographic Institute (NIDI)
Juho Härkönen, Stockholm University

Previous research has shown that parental background is an important predictor of union dissolution, with the intergenerational transmission of divorce gaining the most attention. Next to parental divorce is parental socio-economic status (SES) also shown to be an important determinant for union dissolution decisions. Although the findings from previous studies are not consistent, most existing studies show that the higher the status of the parents, the more likely individuals are to dissolve their own union. However, all these studies are conducted in a single country, while it can be expected that the strength of the impact of parental status on union dissolution depends on the societal context and the specific opportunities this contexts offers. Therefore, the current study aims at better understanding the link between parental socio-economic status (including both the educational and the occupational level of parents) and the risk on union dissolution from a cross-national comparative perspective. First, we test whether there is cross-national variation in the link between parental SES and union dissolution. Second, attention is paid to the mediating role played by parental divorce and own educational attainment. Third, we analyze possible country level indicators which might explain this cross-national variation in the impact of parental SES. In this study we focus on two country level indicators; cultural segregation in the intergenerational transmission of liberal values and the level of income inequality within a country. In this study we examine the dissolution from first union, irrespective of whether this was a marriage of unmarried cohabitation. The data used is from the Generations and Gender Programme (GGP). The GGP is a set of national Generations and Gender Surveys (GGS) consisting of 19 countries. Multilevel discrete time hazard models are estimated to incorporate both individual and contextual level factors into the explanatory model.