Carework

Understanding the Paid Care Workforce

Paid care worker at a table with an elderly person

Paid care work forms the backbone of the U.S. health and social care systems, encompassing a wide range of occupations responsible for delivering clinical care, personal assistance, and preventive services. Despite its central role in population health and economic functioning, paid care work remains challenging to measure systematically. Care jobs are often distributed across diverse settings, span multiple occupational classifications, and involve complex combinations of clinical, relational, and administrative tasks that are not always well captured in traditional labor statistics. As a result, researchers require data infrastructure that links detailed employment information with demographic and contextual characteristics to fully understand the paid care workforce.

IPUMS provides harmonized data from major national surveys that make it possible to study paid care work consistently across time and across data sources. Drawing on datasets such as the American Community Survey and the Current Population Survey, IPUMS enables researchers to identify paid care workers using detailed occupation and industry codes and to analyze their wages, hours, job stability, and demographic characteristics. By supporting longitudinal and cross-sectional analyses, IPUMS allows researchers to examine employment trends, workforce composition, and inequality within the paid care sector.

Measuring Paid Care Work

Paid care workers can be identified in datasets like the American Community Survey (ACS) and the Current Population Survey (CPS), available through IPUMS USA and IPUMS CPS. These datasets include detailed occupation codes, which are based on the Standard Occupational Classification (SOC) system, which allow researchers to examine groups such as nursing assistants, personal care aides, home health workers, and community health educators. Industry codes, which are based on the North American Industry Classification System (NAICS),  allow researchers to identify the setting in which an individual is working, such as a hospital, a physician’s office, or a home-based setting. 

By limiting one’s data file to individuals with a specific set of occupation and/or industry codes and drawing on wage, hours worked, and demographic information, researchers can analyze important dimensions of the healthcare workforce such as job quality, employment trends, and disparities across race, gender, and geography. Because the data are harmonized across time, users can trace how care work has evolved in response to demographic shifts, policy changes, and public health crises.

Analytical considerations

Change over time. An important consideration in measuring paid care work is that occupation and industry codes are revised and change over time. They are revised appropriately every ten years due to changes in the types of work that are performed and to enumerate new and emerging occupations. Consequently, to measure occupation and industry over time, researchers either need to standardize occupation codes using crosswalks or use a standardized occupation variable provided by IPUMS. 

Measuring turnover and employment transitions. A question that a lot of researchers, practitioners,and policy makers want to know about paid care workers is how many, or what percent, of workers are leaving their jobs, how many are entering, and how many are changing industries or occupations. Good news! IPUMS CPS includes a harmonized longitudinal panel that allows researchers to track individuals over a 16 month period. When selected into the CPS sample, household members are surveyed in four consecutive months, left un-enumerated during the subsequent eight months, and then resurveyed in each of another four consecutive months. New rotation groups are brought into the CPS sample each month. An employment transition can be identified when an individual was employed as a paid care worker in the prior month and then reports working in a new occupation in the subsequent month. 

Choosing the right dataset. Choosing the right IPUMS dataset to measure paid care workers depends on your research question. IPUMS USA, which includes the American Community Survey, has the largest sample sizes and best representation of workers across occupation, industry, and other factors, including geography, The IPUMS CPS has a smaller sample size, but the longitudinal panel within the CPS allows workers to better measure employment transitions or turnover (described above). Finally, the CPS March Supplement, called the Annual Social and Economic Survey (ASEC), provides more contextual data for researchers, including jobs and income held in the last year and other work characteristics, like health insurance. 

Example studies using IPUMS data

Azaroff LS, Woolhandler S, Touw S, Bor D, Himmelstein DU. Deporting Immigrants May Further Shrink the Health Care Workforce. JAMA. 2025;333(22):2018–2020. doi:10.1001/jama.2025.3544

Baughman RA, Stanley B, Smith KE. Second job holding among direct care workers and nurses: implications for COVID-19 transmission in long-term care. Medical Care Research and Review. 2022 Feb;79(1):151-60. https://doi.org/10.1177/1077558720974129

 

Dill JS, Frogner BK. The gender wage gap among health care workers across educational and occupational groups. Health Affairs Scholar. 2024 Jan;2(1):qxad090. https://doi.org/10.1093/haschl/qxad090

Date
2026-05-26

Other Aging and Health Posts

Understanding Unpaid Caregiving using IPUMS

Family caregiver in a kitchen

ATUS includes two complementary measures of adult caregiving. The first is activity-based, identifying respondents who report specific adult caregiving activities during the diary day. The second is survey-based, identifying respondents who regularly provide unpaid care to an adult who requires assistance. Using these measures either independently or in combination allows researchers to create a fuller picture of caregiving intensity, regularity, and overlap with paid employment or other ways of spending time (e.g., health behaviors, housework, leisure).

The ATUS also collects detailed information on labor market characteristics, such as occupation, wages, and hours worked, which allow researchers to connect caregiving behavior to engagement in the labor force. This linkage makes it possible to study associations between caregiving and employment, including reduced work hours or exits from the labor force—insights that are essential for designing family-friendly policies and supporting caregivers.

Analytical considerations

The ATUS is a one-day snapshot of time use for each individual respondent. Aggregated at the population level across the year, time diaries can be used to generate estimates of the average amount of time per day that respondents spent in adult caregiving activities in a given year. 

 

Researchers who are looking carefully at the data, however, may notice a disconnect for some ATUS respondents between the activities they report in their time diary (first measure of caregiving) and self-identifying as a caregiver (second measure of caregiving). This isn’t an error, but instead reflects reality for some caregivers. For example, a caregiver may share caregiving responsibilities for a parent with a sibling, dividing responsibilities by day. For the caregiver with Monday, Wednesday, Friday responsibilities who provides the activities they performed on a Tuesday, we might not observe any caregiving activities. This doesn’t mean the ATUS respondent isn’t a caregiver; rather they just weren’t surveyed about a day when they performed their typical caregiving activities. 

 

What this means for researchers using the ATUS to study caregiving is that they need to think carefully about who is in their sample and what this means for any conclusions they draw from their analyses. Be clear about defining terms and clearly describing who is and who is not in your analysis. 

Example studies using IPUMS data

Muench U, Jura M, Spetz J, Mathison R, Herrington C. Financial Vulnerability and Worker Well-Being: A Comparison of Long-Term Services and Supports Workers With Other Health Workers. Medical Care Research and Review. 2020;78(5):607-615. doi:10.1177/1077558720930131

Ulrike Muench, Joanne Spetz, Matthew Jura, Charlene Harrington, Racial Disparities in Financial Security, Work and Leisure Activities, and Quality of Life Among the Direct Care Workforce, The Gerontologist, Volume 61, Issue 6, September 2021, Pages 838–850, https://doi.org/10.1093/geront/gnaa190

Ahmed, T., Floro, M.S. Unpaid Care to Older Persons and Tradeoffs in Time Use: The Experience of Working-Age Women and Men in the US. J Fam Econ Iss 45, 71–87 (2024). https://doi.org/10.1007/s10834-023-09890-3

Wiersma Strauss, A. The earned income tax credit (EITC) and time spent helping and caring for adults. Rev Econ Household (2024). https://doi.org/10.1007/s11150-024-09731-8

Erin Ice, Bringing Family Demography Back In: A Life Course Approach to the Gender Gap in Caregiving in the United States, Social Forces, Volume 101, Issue 3, March 2023, Pages 1143–1170, https://doi.org/10.1093/sf/soac041

Date
2026-05-26

Other Aging and Health Posts