Essays in Pension and Long-Term Care Policy

Abstract

This dissertation studies the relationship of informal elder care and the pension system. The thesis consists of four chapters that apply several micro-econometric methods to survey data sets. The first three chapters use quasi-experimental settings to access important margins in the relationship between informal care giving and retirement and labor market behavior. The fourth chapter builds and estimates a dynamic structural model to simulate effects of future reforms to pension and long-term care policy.

The first chapter analyzes the impact of a reduction in women's labor supply through retirement on their informal care provision. Using data from the German Socio-oeconomic Panel (SOEP) from the years 2001- 2016 the analysis addresses fundamental endogeneity problems by applying a fuzzy regression discontinuity design. We exploit early retirement thresholds for women in the German pension system as instruments for their retirement decision. We find significant positive effects on informal care provided by women retiring from employment at the intensive and extensive margin that are robust to various sensitivity checks. Women retiring from full-time employment, highly educated women and women providing care within the household react slightly stronger. Findings are consistent with previous evidence and underlying behavioral mechanisms. They point to a time-conflict between labor supply and informal care before retirement. Policy implications are far-reaching in light of population aging. Prevalent pension reforms that aim to increase life-cycle labor supply threaten to reduce informal care provision by women and to aggravate the existing excess demand for informal care.

The second chapter examines the effect of an increase in the early retirement age (ERA) for German women on their informal care activity, assessing whether a time conflict between informal care activities and labor supply exist before retirement benefits can be collected. The 1999 pension reform abolished the ERA at age 60 for women born from 1952 onward and therefore supplies quasi-experimental exogenous variation in retirement behavior. I first estimate reform effects on informal care supply applying a regression discontinuity design. Then the reform is used as an instrument for retirement to estimate an elasticity parameter. I apply SOEP data and find that affected women decrease their non-intensive care activity due to the reform and further, my results support the notion that retirement indeed has a causal impact on informal care provision. In a heterogeneity analysis I show that the group of women that is more attached to the labor market reacts more strongly.

In the third chapter we estimate the effect of unemployment on informal care provision. For the identification we use plant closures as a source of exogenous variation and apply a difference-in-differences matching estimation procedure. The analysis is based on data from the German Socio-Economic Panel (SOEP). We find that there is a time conflict between employment and informal care provision. Unemployment increases the probability of providing care by 2.9 percentage points while the daily hours of care provision rise by around 0.05 hours per week-day. In the heterogeneity analysis we show that both men and women react with significant increases in care provision and we find the largest effects for women with low education.

In the fourth chapter we develop a comprehensive life-cycle model of elder parent care and work to evaluate options that address pressing conflicts between pension and long-term care (LTC) policies. Many OECD countries react to challenges of demographic change by increasing LTC by family members (informal care) and raising retirement ages. This intensifies conflicts between paid employment and informal care provision. We extend the previous literature, integrating formal and informal care options to point to impacts of institutionalized incentives on the care-mix. We combine endogenous with exogenous processes and improve on earlier models by incorporating important information on parents to model care-demand. We validate the model using a quasi-experimental setting in Germany. Policy simulations show a decrease in informal care supply as retirement ages are increased. Even though formal and informal care are no perfect substitutes in the model, the demand for formal care increases as a consequence. Further, women with potential care-demand suffer higher reductions welfare. Policy simulations suggest that pension points collected in times of informal care supply reduce detrimental effects of changes to pension rules on informal care supply and the care-mix. These policies can also reduce losses in welfare for women with potential care-demand. Labor market frictions matter in the uptake of informal care. Our simulations show that removing these have similar positive effects on the care system while reducing labor supply.