Partnership transformation in three former socialist countries: Romania, Bulgaria and Hungary. Effects of non-marital childbearing

Mihaela Hărăguş, Babes-Bolyai University

The existent research documents that in the eastern European countries the non-marital living arrangements (single or in cohabitation) are transformed before or soon after the birth of the child, because marriage continues to be the preferred context for childbearing and childrearing. The aim of this paper is to study the transformation of consensual unions and the role of conception/childbirth in this process in three neighbouring former socialist countries: Romania, Bulgaria, Hungary. We conduct our analysis on data from Generations and Gender Survey and our focus is on first time cohabiting women, and our dependent variable is the (first) partnership transformation, which could be marriage or separation, treated as two competing risks. We construct piecewise constant exponential event history models, where the baseline hazard is duration of cohabitation in months. Our main covariate is the pregnancy-and-parity status (time-varying), while other factors that are controlled in the model are age at union formation, calendar period, current educational attainment, living arrangement during childhood. Results for Romania show that pregnancy increases the risk to make the transition to marriage, while once the child is born the risks of transition decrease, compared with not being pregnant. Consensual unions formed at early ages are more likely to transform into marriages, while those formed above age 30 are less likely, compared with unions formed at 23-26 years. The normative pressure to legitimate the union has been continuously decreasing after the change of the political regime in 1989. Regarding the dissolution of cohabitation, conceiving and giving birth to a child while in consensual union increase the risk of separation, compared with non pregnant women, but this effect exists only when the child is very small (below one year). Having conceived or born the child prior to union formation does not show an effect on the risk of separation.