Who Plays the Role of Dad? Intergenerational Transmission in Stepfamilies

Suzanne de Leeuw, University of Amsterdam
Matthijs Kalmijn, University of Amsterdam

Classic mobility research taught us that the resources of biological parents positively affect the educational success of children in traditional two-parent families. Only recently the transmission process in the growing number of non-intact families gained attention. In this type of research the role of non-residential parents and stepparents in the transmission process is examined. The results show that stepparents are also able to influence the educational success of their stepchild and that non-residential fathers might even lose their influence if a stepparent enters the household. However, these studies struggle with data limitations. Recently, a research team at the University of Amsterdam conducted together with Statistics Netherlands a new multi-actor survey among adult children (age 25-45) in the Netherlands (N=6,458) (OKiN data). We use these new data to compare children who experienced a parental divorce and lived in stepfamily during childhood (N=1,808) to children from intact families (N=2,048). Using these data we have two advantage compared to previous studies. First, we use three indicators of resources in the family: educational attainment, occupational status and cultural capital and we are not limited to only educational attainment or occupational status. Second, we are able to take measures of contact frequency and parental involvement into account which was impossible in previous studies. We have all measures of resources and involvement separately for each parent type (residential/non-residential/step). We start with a systematic analysis of the degree to which biological fathers and stepfathers resemble each other and the role that homogamy plays in this resemblance. Subsequently, we use interactions in multivariate regression analyses to compare the influence of biological fathers and stepfathers, controlled for the resources of the mother, on the educational attainment of the child. In addition, we control for parental involvement and contact frequency.

Extended abstract

Due to the increasingprevalence of divorce, more and more children in Western countries are growingup in non-intact families. Some of these children live in a single-parenthousehold but it is relatively common that parents enter a new relationship at somepoint in time; of all children that experience a parental separation in Europe,between 14% (Georgia and Italy) and 60% (Belgium) lives with a stepparentwithin six years after the separation (Andersson,Thomson, & Duntava, 2016).  Classic mobilityresearch taught us that the resources of biological parents, for examplemeasured by the educational attainment or cultural capital of parents,positively affect the educational success of children. However, this type ofresearch is generally based on the assumption that children grow up intraditional two-parent families. Only recently the transmission process innon-intact family structures gained attention (Eriksen,Sundet, & Tambs, 2013; Erola & Jalovaara, 2017;Kalmijn, 2015). These studies examine the role of non-residential parents andstepparents in the transmission process. All existing studies find thatstepparents are able to influence the educational attainment of theirstepchild. One study shows also that the influence of biological fathers disappearsalmost completely when a stepfather enters the household. In other words:“Fathers are replaceable” (Erola & Jalovaara, 2017:971).

Although the existingstudies are a valuable contribution the literature, they struggle with datalimitations. Some use register data which ensures a sufficient number of casesbut limits the analysis to a small and rather basic set of variables. Othersuse survey data which gives access to a wider variety of variables but hasoften a low number of respondents in non-intact families. Recently, a researchteam at the University of Amsterdam, together with Statistics Netherlands,collected a new multi-actor survey among adult children (23-45) (anchor), theirparents and the current partners of their parents (if divorced or widowed)(alter) (Kalmijn et al., 2017) (OKiN). This hasresulted in two datasets: An anchor dataset which contains information on 6,485adult children and an alter dataset which consists of 9,325 (step)parents. Wemainly use the anchor dataset in this research. Non-intact families and inparticular stepfamilies are systematically oversampled via the registers. Forthe purpose of this study we select 1,808 respondents who experienced aparental divorce before the age of 18 after which they lived with their motheror both parents (co-parenting) and in which the mother had a new partner(stepfather) during their childhood. These families are compared to intactfamilies in which the biological father and mother stayed together duringchildhood (N=2,048).  

The OKiNdata have two major advantages compared to data used in previous studies as aresult of which we are able to improve upon existing research. First, previousstudies were limited to the occupational status or educational attainment ofparents as an indicator of resources in the family while we are able examinethe influence of educational attainment, occupational status and culturalcapital of all parent figures (including step) on the educational attainment ofthe child (De Graaf, De Graaf,& Kraaykamp, 2000). In addition, we have data onfinancial resources at the household level (not person-specific). Second, manystudies suggest that one of the central mechanisms of transmission process isthe level of parental involvement (Coleman 1988, Teachman, Paasch and Carver 1997). In contrast to previous studies, we have awide set of variables to examine the degree of parental involvement of allthree parent types empirically. We use two direct measures of parentalinvolvement. First, a scale with items indicating whether parents help withhomework, talk about personal and school related matters with their child,brought them to sport activities and had fun day outs together with theirchild. Second, an item measuring how influential a parent figure was whenimportant decisions in the life of a child had to be made. In addition, we knowhow much time parents and children lived together and how much contact therewas with a non-residential parent.

We start with asystematical examination of the degree to which fathers and stepfathersresemble each other and the role that educational homogamy plays in generatingthis resemblance (Gelissen 2003).  Next, we use multivariate regressionsto answer our research questions. Interactions are used to examine theinfluence of the father and stepfather, controlled for the resources of themother, on the educational attainment of the child. In the subsequent models wewill add controls for the level of parental involvement and contact. In thisway we get a better understanding of the results of the few previous studiesthat looked into the intergenerational transmission process in stepfamilies.

Presented in Session 1114: Families and Households