Essays on Three Social Insurance Programs: Design & Consequences for Work and Family Decisions

Albanesi (2016) defines social insurance schemes as government transfer programs whereby individuals who claim a condition or state that reduces their labor income obtain a transfer from the government. Which "conditions'' and "states'' should be covered by social insurance, as well as the optimal level of coverage, has been the source of debates among policy-makers and economists for more than a century.Early on, working women fought for protection against income loss resulting from childbirth. Back in October 1919, delegates of the first International Congress of Working Women gathered in Washington D.C. and successfully convinced the International Labour Organization to adopt a maternity protection convention, which mandated that maternity leave must come with benefits "sufficient for the full and healthy maintenance'' of mother and child. Several decades later, in the 1970's, Scandinavian countries adopted gender-neutral parental leave policies, which opened government paid leaves to fathers. They have since been followed by more than 90 countries which offer paid leave mandates to both parents. Probably the largest outlier among non-adopters of parental leave policies is the United States, whose 46th president Joe Biden failed during the first year of his mandate to fulfill his campaign promise to put in place a nation-wide parental leave program. The century-long reforms across the world, as well as the ongoing debate in the USA, one of the largest labor markets, motivates a deeper understanding of the consequences of parental leave policies for work and family decisions.The first two chapters of this thesis find their origin in previous work, which studies the labor market trajectory of Belgian women around childbirth. This earlier study demonstrates that Belgian mothers lose about 30 percent of their labor earnings relative to fathers, up to eight years after the birth of their first child (Fontenay et al. 2021). Figure I at the end of the introduction illustrates this dynamic and has been the common thread of the first two chapters. Several questions can be traced back to this figure: Why do women already experience an income loss when they are pregnant with their first child? How do women react to the income loss resulting from the incomplete wage replacement offered by the maternity allowance after childbirth? What are the drivers of the persistent drop in wage income up to eight years after the birth of their first child? This thesis tries to offer an answer to all of these questions using empirical methods in the context of Belgium.From a conceptual point of view, this thesis also makes the case that social insurance programs should not be considered in isolation because they can have important spillover effects. After reading this thesis, the reader should find that social insurance programs shape work and family decisions, but also that changes to one program might have unintended consequences for other branches of social security. From a methodological point of view, the three chapters rely on modern econometric methods for policy evaluation. This work uses both natural experiments and a randomized controlled trial to evaluate the effects of policies and their consequences for workers.The first chapter originates from the observation that the level of compensation during maternity leave varies significantly across countries, but previous research on the topic provides only limited insights on the consequences of this important design choice. More particularly, the first chapter assesses how the generosity of maternity leave allowance affects first-time mothers' career trajectory and subsequent fertility decisions. It exploits the fact that the allowance is capped in Belgium, so that women with pre-leave earnings above the maximum threshold face drastically lower replacement rates. Using a regression kink design, as well as a rich set of administrative data on mothers from 2002 to 2015, this chapter highlights the consequences for their career and show that mothers who receive higher benefits are more likely to leave salaried employment for self-employment. The study also reveals that mothers who receive a more generous allowance have more children and it provides suggestive evidence connecting this fertility effect to the transition to self-employment.The second chapter, co-authored with Ilan Tojerow, explores the short-run and long-run determinants of the income loss suffered by young mothers. The study shows that a woman’s likelihood of claiming disability insurance increases after the birth of her first child, but also reveals that the provision of paternity leave can ease this effect. Using Belgian administrative data, the findings are twofold: the incidence rate of disability is constant across gender up until a woman becomes a mother for the first time, and the provision of paternity leave, even in short intervals, significantly reduces the number of days that mothers spend in disability. Specifically, the regression discontinuity difference-in-differences design shows that mothers with partners eligible for two-week-long paternity leave spend on average 21% fewer days in disability over twelve years. The fiscal consequences of the research suggest that spending on paternity leave would be more than compensated by the savings in mothers’ disability benefits.The second chapter also highlights the increasing number of workers who rely on another social insurance program, called Disability Insurance (DI), as they find themselves unable to work because of a medical condition. A large share of these beneficiaries enters DI because of mental health conditions, which generally start at a young age and too often translate into a lifetime of claiming benefits.The third chapter, also co-authored with Ilan Tojerow, evaluates the effects of a Supported Employment (SE) program aimed at helping DI recipients with mental conditions to return to work. The program is characterized by a "work first'' approach with intensive job counselling and follow-along support. Using a Randomized Control Trial with more than 660 participants over a follow-up period of 18 months, the study compares the benefits of this newly introduced program to regular vocational rehabilitation services traditionally used in Belgium. The results show that SE increases the probability of DI recipients with mental conditions to work while on claim and reduces their reliance on DI benefits. Specifically, the estimates reveal that, 18 months after the start of their return-to-work program, participants in the SE group are 9.6 percentage points more likely to be working and receive 6.5% less in DI benefits than those in the control group. The effects of SE remain substantial even for those participants who were impacted by the COVID-19 pandemic. Cost-benefit analysis suggests that spending on SE could be compensated within less than two years by the savings in DI benefits.Some important policy implications can be drawn from these studies regarding the career of women and the effects of public policies. First, providing young mothers with more generous maternity leave allowance would help reducing the gender gap in entrepreneurship, while having positive spillover effects on fertility. Second, paternity leave policies, by increasing the involvement of fathers and co-parents, may improve the long-term health of mothers and prevents career breaks due to sickness or disability. Third, for those who are already on disability insurance, supported employment policies may help them return to work and avoid a lifetime of claiming benefits. All these findings are also of particular importance for the economy as a whole since previous research has revealed that a better allocation of women's talent would boost productivity and growth (Andrew, Bandiera, Costa-Dias, & Landais, 2021; Hsieh, Hurst, Jones, & Klenow, 2019).