Repartnering of Women: The Role of Economic Hardship and Vulnerability
Katya Ivanova, University of Amsterdam
Nicoletta Balbo, Bocconi University
Alessandro Di Nallo, Universitat Pompeu Fabra
There are relevant implications for investigating the process of repartnering. According to McLanahan (2004), in the US higher-educated women tend to have gaining trajectories, with later childbearing, maternal employment and lower risk of union dissolution, while women with fewest resources tend to pursue loosing trajectories with early, non-marital childbearing and higher risk of separation. There has thus been a divergence in the opportunities for these groups of women, and an increasing disparity of life chances for their children. A separation has a harsh influence on economic wellbeing of both ex-partners (Amato, 2010). Further, it “is a stratifying event” (Shafer, 2013), as it hits more hardly the individuals with fewest resources, and influences women’s well-being more negatively than men’s (McManus & Di Prete, 2001). Compared to men, women have on average less work experience and earnings, but bear more caregiving responsibilities after a union dissolution (Smock et al., 1999). The impossibility of relying on a partner’s income, along with the financial drain caused by children’s maintenance, is likely to take a toll on low-income women.
Existing research on re-partnering has highlighted several factors involved in new union formation: age, gender, personality, socio-economic status and parental status determine why some categories of separated are more likely to find a new partner than others (e.g., Shafer, 2009; Shafer, 2013; Gelissen, 2004). Need, attractiveness and opportunity are the three most cited channels that influence the patterns of repartnering after a union dissolution (de Graaf & Kalmijn, 2003). Need, in particular, is usually deemed as the spur to overcome singlehood, achieve companionship (Amato, 2000) and escape from economic hardship (Jansen et al., 2009), especially for women (Poortman, 2010).
A growing strand of literature is investigating the question “Who is repartnering whom?”, namely which kind of new partner the people form a union with. The few studies who address this issue (e.g., Shafer, 2009, 2013) do limit their span of research only to the formerly married and to those who form a new marriage. However, evidence from the US (Wu & Schimmele, 2005) highlights a decreasing probability of marriage as a relationship setting for the newly repartnered and find an increasing proportion of new union formations among the cohabitations. Further, little investigation in the US is devoted to understanding whether low and high SES women do find systematically different partners, and whether these changes vary according to their parental status and the economic cycle.
In this article, we study the patterns of socio-economic assortative mating after a marital or cohabiting union dissolution in the United States using data from the National Longitudinal Survey of Young Women 1968, the National Longitudinal Study of Youth 1979 and 1997, which follow 4,612, 6,592, and 5,159 women, respectively. In our work we address the following questions: 1) What is the influence of women’s SES on repartnering? 2) What is the influence of women’s own SES on which partners they repartner with (in terms of SES)? 3) Does the SES of the ex-partner matter for which partner women repartner with?
Using discrete time event history analysis to model the process of re-partnering, we estimate three models. The first model assesses the impact of the respondent’s SES (social class or educational attainment) on the probability of repartnering on a given month with a logistic regression. The second model is a competing risk model and estimates the probability of repartnering with a person of a specific SES (vs. remaining single), given that one has not repartnered yet. The third model is a competing risk model whose goal is to take into account the “educational match” in the previous union and check whether women tend to repeat their repartnering pattern across time.
A group of independent variables are assumed to proxy for conditions of economic hardship: parental status (being mother vs. childless) and economic crisis. We hypothesize that specific conditions of “need” – for instance, the separated woman having dependent children or the negative economic cycle – change the patterns of assortative mating between the low and the high SES women. In particular, low SES women might be more in “need” when taking on the post-separation custody of their children compared to their high SES counterparts and to childless women. We explore different research designs to assess the moderator role of the economic cycle: (a) we intend to employ a time-varying measure such as the regional or nationwide unemployment rate or GDP growth; or (b) carry out distinct models in periods of economic turmoil (1973-1974 or 2007-2009) vs. economic growth.
Presented in Session 1111: Families and Households