Continuities in antisocial behavior and parenting across three generations


Carolyn A. Smith, School of Social Welfare, University at Albany, State University of New York, 135 Western Avenue, Albany, NY 12222, USA; Email:


Background:  Accumulating evidence indicates that there are intergenerational continuities in antisocial behavior, and that parenting patterns play a role in these continuities. Very few studies, however, enable assessment across two generations of children at comparable ages, employing independent reporters and comparable measurements. The present study addresses the extent to which antisocial behavior in parents predicts antisocial behavior in children in two successive generations; the degree to which a man's childhood antisocial behavior predicts antisocial behavior in his own children; the extent to which parenting problems are related to child antisocial behavior similarly in two successive generations; and the extent to which intergenerational continuities in antisocial behavior are mediated by parenting variables.

Methods:  Questions are addressed with prospective longitudinal data from the Cambridge Study in Delinquent Development (CSDD). The CSDD includes data on 411 Inner London males (Generation 2, or G2), their female partners, their parents (G1) and their children (G3). At time 1, when G2 were aged 8–10, data on G2 child conduct problems and G1 parenting and convictions were available. At time 2 when G2 males were aged 32, data were available on the parenting of 178 G2 fathers with G3 children aged 3–15, and on their G3 children's behavior. Time 1 data come predominantly from G1 mothers, whereas time 2 data come predominantly from G2 fathers.

Results:  Between generations, antisocial G1 mothers and fathers predicted conduct problems in G2 and G3 children, but G2 child conduct problems did not predict G3 child conduct problems. Within generations, G2 child conduct problems predicted G2 adult antisocial behavior and antisocial partnerships, which in turn predicted G3 conduct problems. Parental conflict and authoritarian parenting were similarly related to early childhood conduct problems in two successive generations. There was relatively little continuity between G1 and G2 parenting except that G2 males who were poorly supervised by their parents were themselves poor supervisors as fathers. Both G1 and G2 generations displayed assortative mating, with antisocial males tending to partner antisocial female peers.

Conclusions:  There are between-generation and within-generation continuities in antisocial behavior, although assessment of such continuities is complicated by inevitable design and measurement limitations. Parenting partly mediated the impact of parental antisocial behavior on child antisocial behavior in two successive generations, but the relation between antisocial parents and antisocial children is not fully mediated by parenting variables.