Care Giving and Labor Market Outcomes
Paper Session
Sunday, Jan. 3, 2027 8:00 AM - 10:00 AM (EST)
- Chair: Usha Nair-Reichert, Georgia Institute of Technology
Remote Work, Occupation Sorting, and the Motherhood Penalty
Abstract
The motherhood earnings penalty persists most strongly in occupations characterized by long hours and high face-time demands—business, finance, and law—yet many of these occupations are technologically teleworkable. I term this mismatch "Latent Remote": occupations teleworkable in principle but office-bound by face-time requirements before COVID. The pandemic relaxed this constraint, offering a natural experiment to identify how locationflexibility—distinct from the schedule flexibility in Goldin (2014)—reshapes the motherhood penalty.
Using panel data from the Survey of Income and Program Participation (SIPP) around first childbirth, I estimate post-birth outcome trajectories by pre-birth occupation teleworkability via the pseudo-panel event study of Kleven et al. (2025). Mothers in high-teleworkable occupations experience roughly half the post-birth employment decline of mothers in low-teleworkable occupations and recover earnings substantially faster. The divergence operates through extensive and intensive margins of labor supply rather than hourly wages, and high-teleworkable mothers shift sharply into remote work post-birth.
To quantify mechanisms, I develop and estimate a dynamic discrete choice model of labor supply, occupation choice, and remote work take-up over the life cycle, with multidimensional occupations characterized by task portability and face-time demand. The model decomposes the motherhood penalty into two channels: human capital accumulation through labor force participation, and occupation sorting around childbirth. I model the post-COVID rise
in remote work as a demand-side expansion of the choice set. Counterfactual simulations quantify how this expansion reshapes each channel.
The findings establish location flexibility as a central, previously underappreciated determinant of the motherhood penalty, particularly for women in high-skill occupations long viewed as structurally family-incompatible.
The Hidden Costs of Care: Parental Dementia and its Impact on Work and Retirement
Abstract
As dementia prevalence rises and long-term care needs grow, much of this care is provided informally by family members, often adult children who are approaching retirement. This paper examines how caring for a parent with dementia affects adult children’s labor supply and retirement decisions, with attention to differences by marital status and gender. This question is important because dementia caregiving often occurs during the late working years, when reduced labor supply can have lasting consequences for retirement security.Using RAND HRS and family data linked to Social Security administrative records, I show that dementia caregiving rises through the 50s and early 60s, Caregiving hours are substantially higher when parents have dementia, especially for women and single individuals. Within married households, descriptive evidence suggests specialization and an unequal division of care: married women provide more care than married men; the caregiving spouse works fewer hours and is more likely to rely on spousal benefits, while the other spouse works more and is more likely to claim on own earnings records.
Using linear probability models with rich socioeconomic controls and survey-wave fixed effects, I find that dementia caregiving is associated with lower labor supply and earlier benefit claiming. Among married individuals, caregiving raises the probability of claiming benefits by 10–15 p.p. at ages 63–64 and lowers the probability of full-time work by 10 p.p. at ages 62–63. These effects are particularly strong for married women, while the estimates for married men are not statistically significant. Single caregivers likewise reduce work and claim benefits earlier.
The findings suggest that dementia caregiving imposes meaningful economic costs that are unevenly distributed across households and highlight within-household coordination among married couples. By highlighting how caregiving interacts with Social Security, the paper informs retirement and caregiving policies, including caregiving credits to reduce penalties faced by family caregivers.
Is It Caring Labor, or Is It Women's Work? Objective Classification, Perceived Norms, and Gender Segregation
Abstract
Gender segregation in caring labor occupations remains persistent and poorly understood, in part because of a fundamental measurement problem: existing definitions of caring labor conflate task content with gender composition, making it impossible to cleanly separate whether women sort into caring occupations because of what those occupations involve, or whether occupations are perceived as caring because women dominate them.This paper addresses the circularity by separating two classifications that have historically been entangled. I use large language models (LLMs) applied to anonymized occupation task-content descriptions to construct an objective, content-based classification of caring labor, scored on five theoretically grounded dimensions from the foundational literature (Anker 1997; Folbre 1995; England 2005): direct human interaction, emotional labor, nurturing content, relational work with vulnerable populations, and service orientation toward others' well-being. I compare this LLM-derived classification against a subjective classification elicited from a survey experiment of 1,000 respondents rating 177 occupational categories on a five-point caring-labor scale. The gap between the two captures the gender-stereotype component of how caring labor is socially constructed.
I show that the gap is structured: it is larger for female-dominated occupations, consistent with stereotype-driven perception beyond what task content predicts. I further decompose the gap across four norm dimensions: worker-side beliefs about women's suitability, employer-side hiring preferences, recipient-side consumer preferences, and cross-gender transgression norms measuring the social cost of gender-atypical care. Respondents with more traditional gender-role attitudes show larger gaps, and the transgression items reveal a gender-asymmetric pattern: tolerance for women crossing into male care spaces is systematically higher than for men crossing into female ones.
Finally, I link the LLM-based classification to women's occupational sorting in the China Family Panel Studies, decomposing the norms-to-occupation channel into content-based sorting and stereotype-driven perception. This decomposition has direct implications for where policy interventions targeting caring labor segregation should be aimed.
Who Earns, Who Cares: Earnings Power, Gender, and the Allocation of Labor Supply and Caregiving Within Older Couples
Abstract
This paper asks how married older adults divide caregiving across three recipients — parents and in-laws, grandchildren, and a disabled spouse — and how this division shapes labor supply. We focus on whether earnings power within the couple reorganizes care across recipients and whether it operates symmetrically for men and women.Using the Health and Retirement Study (HRS) Core, the HRS Family supplement, and the Consumption and Activities Mail Survey (CAMS), we show that women provide more care than men in every setting and bear the cost through reduced full-time work rather than by leaving the labor force. In the full married sample, primary earners provide 4–5 percentage points (pp) less parental care than non-primary earners, with no statistically significant effect on grandchild care. Among households where the spouse has a daily-living limitation, primary earners are 13.5 pp more likely to provide spousal care. The two forces interact unequally by gender. Female primary earners do not reduce parental care the way male primary earners do, and they keep working full-time when they take on caregiving, while non-primary-earner women shift to part-time. For spousal care, male primary earners both work more and provide more care than other men: they supply 4.68 more weekly hours of spousal care and are 17.5 pp more likely to be employed. Female primary earners are only about 3 pp more likely than non-primary-earner women to provide spousal care.
Earnings power within the couples reorganizes spousal care very differently from parental or grandchild care. The findings imply that caregiver support policies should account for both the dual burden of work and care among male primary earners and the reduction in full-time work among non-primary-earner women.
JEL Classifications
- J2 - Demand and Supply of Labor
- J1 - Demographic Economics