Welfare Policy and the Implications of Divorce for Labor Market EarningsWelfare Policy and the Implications of Divorce for Labor Market Earnings

This study examines relations between the development of welfare legislation affecting singleparent families in Israel and the implications of divorce for mothers' earnings. The rise in singleparent families has been one of the most prominent social developments since the last quarter of the twentieth century (Kilkey, 2000; Lewis & Hobson, 1997; Millar & Rowlingson, 2001; Orloff, 2001). The relative number of these among all families with children is growing steadily. In some countries, more than half the children will be part of a single-parent family at one point or another in time (Millar & Rowlingson, 2001). One of the common characteristics of single mothers in many countries is their poverty rates, which are remarkably higher than for the rest of the population. In the vast majority of cases, such families are headed by divorced mothers. There is an abundance of research reporting a significant decline in the economic wellbeing of women (and children) in the post-divorce period (Andreß, Borgloh, Bröckel, Giesselmann & Hummelsheim, 2006), whereas the consequences of divorce on men's economic wellbeing are less severe (Bianchi, Subaiya & Kahn, 1999; Jarvis & Jenkins, 1999; McManus & DiPrete, 2001; Manting & Bouman, 2006). The high rates of divorce and its economic consequences, particularly for women, have given rise to state policies targeted at the single-parent household (Fraser & Gordon, 1994; Stier, 2011), especially with respect to integration into the job market (Abramovitz, 2006; Cook, 2012; Kilkey, 2000; Millar & Ridge, 2008; Orloff, 2001; Skevik, 2005). Welfare practices (e.g., allowances, social services, job market protections) impact the risk of deteriorating into poverty and the ability to maintain autonomous households (Kilkey, 2000; Lewis & Hobson, 1997; Millar & Rowlingson, 2001). Whereas in some welfare states, single mothers have been expected to serve as ‘full-time care-givers' for their children, in others they are also expected to work (Kilkey, 2000). Influenced by a neoliberal ideology, the end of the twentieth century saw the beginning of a global policy of ‘activation', i.e., a gradual attempt to integrate single mothers into the workforce accompanied by welfare-to-work programs and increasingly strict means and employment tests for welfare entitlement (Cook, 2012; Skevik, 2005). Previous studies indicate that, in some countries, welfare solutions strengthen the position of divorced women in the job market. Welfare policy encourages the employment of women by offering a wide range of childcare frameworks, extended maternity leave and a broad public sector providing job opportunities suitable for parents (Mandel & Semyonov, 2005). The welfare policy that encourages women to enter the job market thus improves the economic conditions of their families following a divorce (Raeymaeckers et al., 2008; Uunk, 2004). Following previous studies (McKeever & Wolfinger 2001; Raeymaeckers, Dewilde, Snoeckx & Mortelmans, 2008; Raz-Yurovich, 2013; Van Damme, Kalmijn & Uunk, 2009), the current research focuses on the implications of divorce for earnings of mothers in Israel, in the context of changes in the welfare state. Israel constitutes an important case study because of these changes in its welfare policies coupled with high fertility rates (Fogiel-Bijaoui, 1999; Gal, 2010). In Israel of the 1990s, the economic autonomy of single mothers was established through a combination of allowances and paid labor. However, starting in 2003 the political climate undermined the 30 legitimacy of welfare state-based economic autonomy, and since then the welfare state model has turned towards the market. Following the analytical strategy of Styrc and Matysiak (2012) and of Tamborini, Iams and Reznik (2012), we examine the implications of divorce for mothers' earnings in two periods of time that represent two kinds of policies targeted at single-parent families. While in the first period (1996-2003) economic autonomy of single mothers was established through a combination of allowances and paid labor, in the following period (2004-2008) legitimacy of welfare state-based economic autonomy was undermined, and the welfare state model turned toward the market. The current research suggests a theoretical and empirical meeting of institutional and micro levels of analysis (Härkönen & Dronkers, 2006). Specifically, we examine policy changes affecting singleparent families, analyzing the implications of these changes for the earnings of mothers following divorce. Our research question is: How does divorce affect mothers' earnings, and how is this related to changes in welfare policy between the two periods? Method.We used data from the combined Israeli census files (from the Central Bureau of Statistics; hereafter CBS) for the period 1995–2008. These data were matched to annual work history data and income tax data for each year from the National Insurance Institute (NII), as well as the civil registry of divorce (via the CBS). Annual data for each mother includes information about yearly earnings from paid labor, the number of months employed over the year and the mother's marital status in that year. The other independent variables that appear in the analysis (e.g., duration of marriage, ethnicity) are based on the 1995 census. Our analysis was designed to explore the change in mother's earnings after divorce and how this was related to change in policy configurations across time. Therefore, we restricted the sample to salaried mothers between the ages of 18 and 55 who were married and living with their spouse in the same household at the beginning of the research period (1995) and had divorced between 1996 and 2008 (n=1,466). To examine changes in earnings following divorce, and whether these earnings changed from one policy period to the next, we first present descriptive statistics of median annual earnings and percentage change in annual earnings one year prior to the divorce and one to three years after the divorce in the two policy periods. Second, we utilized two types of multivariate regression techniques: (1) OLS regressions estimating the log difference between mother's pre- and postdivorce (at T+1, T+2 and T+3) earnings; and (2) logistic regressions estimating the log odds that a mother's real post-divorce earnings (at T+1, T+2 and T+3) were at least 30% higher than her earnings at T-1. These analyses should be treated as descriptive regressions (Ginther & Pollak, 2004), as they are not intended to evaluate causal relationships. This kind of analysis furthers understanding of the conditions that differentiate between patterns of change in mothers' postdivorce earnings (for a similar analysis, see Tamborini et al., 2012). Findings. In general, we found that, regardless of policy period, on average divorced salaried mothers tend to increase their earnings one to three years following the divorce, consistent with previous research (Raz-Yurovich, 2013; Tamborini et al., 2012). About 80% of the divorced salaried mothers in our study experienced a post-divorce real earnings gain of at least 10%. However, the ability to increase earnings was higher in the first policy period than in the second. Our multivariate analysis confirmed this connection between welfare policy period and the economic implications of divorce for mothers. When there was a generous welfare policy (1996- 2003) that encouraged a combination of care work and paid labor, mothers' earnings increased in the one to three years following the divorce more so than in the second period (2004-2008), which encouraged paid labor through allowance cuts. Our study also confirms the importance of certain 31 socioeconomic resources and family characteristics for mothers' earnings following divorce (Jansen et al., 2009; Raeymaeckers et al. 2009; Tamborini et al., 2012; Van Damme et al., 2009). The study suggests that the marked infringement on welfare mechanisms that offered autonomy to mothers reduced their ability to enhance their earnings from paid labor after divorce. A previous study, using the same data set, found the likelihood of divorce to be higher among lower socioeconomic groups (Kaplan and Herbst, forthcoming). Elsewhere (Herbst and Kaplan, 2014) we also showed that, throughout the two periods, married mothers earned more than divorced mothers. It should be reiterated that the number of poor single-parent families rose significantly after the welfare reform (Endeweld et al., 2009). All of this suggests it is essential to create policy that helps these mothers improve their earnings after divorce to make up for their lost income (ex-husband's earnings). This kind of policy supports lower socioeconomic groups and can therefore reduce poverty.