Caregiving and Labor Force Participation: New Evidence from the American Time Use Survey

A large share of the growing demand for elder care is met informally by relatives, who provide an estimated $470 billion worth of unpaid care annually (AARP 2013). Understanding the implications of this care, especially in terms of labor supply, is important considering that caregivers are often themselves nearing retirement. Existing studies of caregiving in the US rely primarily on a biennial panel survey, which makes it difficult to closely examine the transition to caregiving and its immediate effects on work behavior and time use. In this paper, we provide new evidence of the short-term impacts of caregiving on labor supply from the American Time Use Survey eldercare module from 2011 to 2016. We link the ATUS to the Current Population Survey (CPS) panel and take advantage of retrospective reports to construct a short (18-21 month) panel, capturing employment outcomes immediately before and after individuals begin providing care. Understanding these short-term impacts is key to the policy debate about how family-friendly workplace policies, such as paid family leave, may help balance the demands of informal care and work.

We look at likelihood of working, hours worked, and absences from work to estimate the impact of caregiving on labor force participation at a much finer interval than existing work. We present evidence on an unexplored dimension of heterogeneity: caregiver education status. Preliminary results suggest that starting caregiving is associated with a 3% increase in stopping work, as well as a 25% increase in the likelihood of being absent from work in a given week. All labor supply changes are concentrated among higher educated (college and above) caregivers. We find little evidence that this is driven by selection into different types (intensity) of caregiving across different groups.