Repartnering after marital dissolution: Does context play a role?

Abstract

Abstract This paper examines in depth the determinants of repartnering in Italy. With data from a national survey conducted in 2003, the effects of socio-economic, demographic and contextual characteristics on second-union formation among separated women are examined. The analysis is of particular interest in a country such as Italy, which is undergoing a transition from traditional to modern family behaviours. In addition, it allows us to verify the hypothesis that the importance of demographic factors in the repartnering process decreases as marital instability becomes more common. Results of event history analysis models confirm the basic hypothesis and show some unexpected effects. 1. Introduction In Italy, the phenomenon of marital disruption is relatively recent and still not very common (ISTAT 2007). This is probably due to cultural and normative factors, which make separation costly, from both emotional and economic viewpoints. In Italy, indeed, the sacredness of the first marriage is still important (Rosina and Fraboni 2004). In addition, divorce was introduced by law only in 1970, and only since 1987 has the time required to request a divorce after legal separation been reduced from 5 to 3 years. Not surprisingly, therefore, in Italy unions after the first marriage are not very common, even when the end of the first union is a choice and is undergone by relatively young individuals: in 2003, couples at their second union, including those after widowhood, represented only 5% of all couples (ISTAT 2006). However, marital dissolution is rapidly increasing (ISTAT 2007), indicating that new unions formed after the first marriage may also be more common. In this paper, we examine the factors influencing repartnering (cohabitation or remarriage) among Italian women who dissolved their first marriage. The international literature is rich in studies on this topic (e.g., Hunt 1966; Thornton 1977; Wu and Schimmele 2005). It has been shown that repartnering may be influenced both by demographic factors (i.e., age of the woman at separation, children born during the first marriage) and by socio-economic characteristics (woman's education and her employment status). To date, however, the literature cannot always explain why some factors (e.g., children, woman's human capital) show mixed effects, and which mechanisms actually explain these results. Our aim is to analyse in depth the determinants of repartnering in order to shed light on the mechanisms underlying this phenomenon among separated women living in a country like Italy, which is undergoing a transition from traditional to modern family behaviours. The few studies about repartnering in Italy are based on data from the second half of the 1990s (Rettaroli 2002; Angeli and De Rose 2003); as the phenomenon was then still recent, these pioneering studies faced limitations in both sample size and variables. The availability of data from the survey Family and Social Subjects, conducted in Italy in 2003, overcame most of these problems and, as a result, the effects of some key factors (i.e., woman's employment status and the characteristics of first-marriage children) may be better analysed. Data from this survey also allowed us to verify whether there is a contextual effect that influences the determinants of repartnering. Our basic hypothesis is that the opportunities of a second union for separated women are influenced by the level of diffusion of marital instability and, more generally, by the level of social acceptance of new family behaviours. In particular, we assume that the social costs of repartnering decrease, passing from a traditional to a more modern context; with them, the importance of some women's demographic characteristics, which influence these costs, also decrease. Differences between the North and South of Italy, according to the diffusion of behaviours typical of the Second Demographic Transition (Gesano et al. 2007; De Rose 1992; Rettaroli 1997), are used to test this hypothesis. …