(Credit: © Wesley/peopleimages.com - stock.adobe.com)
Why Roller Coaster Schedules Could Cost Americans Their Health Insurance
In A Nutshell
- Workers with unstable pay or hours are about 20 percentage points less likely to have job-based health insurance, and those who lost work entirely saw a gap as high as 24 points.
- The pattern held across income swings, hour swings, and job loss alike, even though these workers still averaged more than 30 hours a week when employed.
- A new federal law will require many Medicaid recipients to log 80 hours a month starting no later than January 2027, right as unstable workers routinely cross above and below that threshold.
- Enhanced ACA marketplace subsidies expired after 2025, raising costs for the coverage option unstable workers already lean on more than stable workers do.
Workers whose income or hours were unstable were about 20 percentage points less likely to have health insurance through a job, and those who lost work entirely saw a gap as high as 24 points, according to new research. A retail worker might log 35 hours one week and just 15 the next. That kind of swing can happen without the person leaving the job at all, and it’s common for millions of low- and middle-income workers in America.
A study published in JAMA Network Open looked at what happens to health coverage when paychecks or work hours bounce around unpredictably. The timing matters. A federal law enacted in 2025 will require certain Medicaid recipients to work, train, or do community service for at least 80 hours a month starting no later than January 2027, while enhanced ACA marketplace subsidies expired after 2025. For workers whose hours already swing wildly, both changes could make losing coverage more likely.
That pattern held true across every kind of instability the study measured, and it means employer coverage often isn’t in the cards for these workers, right as the public programs they lean on instead are being tightened.
Unstable Hours Take Three Distinct Forms in the Data
Researchers used a large government survey that follows the same households over time, called the Medical Expenditure Panel Survey. They focused on adults between 18 and 64 years old whose household income was at or below 400% of the federal poverty level and who had worked at some point during 2022 and 2023.
Instead of just asking whether someone had insurance at one point in time, the team measured job instability three ways: swings in yearly wages, how much weekly hours varied round after round, and whether someone had gone from employed to unemployed between check-ins. A person could have wildly swinging hours but never lose the job entirely, while someone else might.
Each unstable-work group was then compared against workers whose employment stayed steady, looking at whether they were offered insurance through their employer and what type of coverage, if any, they ended up with. Researchers adjusted for age, sex, race or ethnicity, marital status, and whether someone had children, so those differences were less likely to explain the insurance gaps they observed.
Job-Based Coverage Gaps Reached 24 Percentage Points
Sample sizes broke down into 943 workers representing about 14.6 million people who experienced income swings, 739 workers representing about 11.4 million people with unstable work hours, and 186 workers representing about 2.9 million people who had gone through a period of unemployment.
Compared with workers in steady jobs, people in all three unstable-work groups tended to be younger, more likely to be female, and had lower incomes overall. Workers across all three groups still averaged at least 30 hours a week whenever they were employed, so these were not people barely working. Their hours simply moved around a lot, often crossing above and below 20 hours a week, roughly the weekly equivalent of the new 80-hour monthly Medicaid requirement. Depending on the type of instability, somewhere between 45% and 83% of these workers had hours that swung above and below that mark, compared with a much smaller share of workers in stable jobs.
Workers with unstable income were about 15 percentage points less likely to even be offered insurance through their job compared with stable workers, and workers with unstable hours saw a nearly identical gap. Gaps in actually having employer coverage were larger still, at about 20 to 24 percentage points across all three groups, and those same workers were between about 12 and 19 percentage points more likely to be covered by Medicaid instead. Workers with unstable income and unstable hours were also somewhat more likely to have ACA marketplace coverage, by about 4 to 5 percentage points, compared with stable workers.
Policy Changes Threaten to Compound the Coverage Gap
These patterns collide with the new Medicaid rules and the loss of enhanced marketplace subsidies. The law requires eligible enrollees to log 80 hours a month of work, training, or community service, with some exemptions, and to prove it through regular eligibility checks. Since unstable workers in this study routinely crossed above and below the roughly 20-hour-a-week mark, even while averaging more than 30 hours a week overall, the authors say that could leave some of them vulnerable to losing coverage over documentation problems rather than an actual lack of work.
Losing the enhanced marketplace subsidies compounds the problem. Unstable workers leaned on marketplace coverage more than stable workers did, so losing that assistance makes one of their main backup options more expensive right when they need it most.
For millions of workers whose schedules bend and snap from one week to the next, that’s the bind: unstable hours were linked to missing out on job-based coverage, yet not to being exempt from new rules built for people with steady schedules. Coverage remains a moving target, right as both policy changes make missing it more costly.
Disclaimer: This article is based on findings from a peer-reviewed observational study and reflects the data and interpretations reported by the study’s authors. It is intended for general informational purposes and is not legal, medical, or financial advice regarding health insurance or public benefit programs.
Paper Notes
Limitations
Authors noted that the study relied on self-reported survey data, which can be less precise than employer or government records. They also flagged that the study period overlapped with the unwinding of continuous Medicaid coverage that had been in place during the COVID-19 pandemic, a shift that could affect how well these findings apply to other time periods. The study is observational, so it shows associations between unstable work and insurance coverage rather than proof that instability directly caused the coverage gaps. It also does not identify why any individual’s hours or income became unstable, or how many workers will ultimately lose Medicaid coverage under the new federal work requirement.
Funding and Disclosures
Authors reported financial relationships outside this specific study. One author disclosed grants from the National Institutes of Health, the Agency for Healthcare Research and Quality, the Michigan Department of Health and Human Services, Blue Cross Blue Shield of Michigan, the Commonwealth Fund, and the Michigan Health Endowment Fund, along with past unpaid board service for the Center for Health and Research Transformation. Another author disclosed grants from the National Institute on Aging, the Michigan Department of Health and Human Services, and the Centers for Medicare and Medicaid Services, as well as personal fees from JAMA Network, Brigham and Women’s Hospital, Weill Cornell Medicine, the University of California Davis, and the Japanese Society of Clinical Epidemiology, plus nonfinancial support from the National Academy of Medicine, Princeton University, and the University of California Los Angeles. A third author disclosed grants from the Doris Duke Foundation, the W.K. Kellogg Foundation, and the Kresge Foundation. No other disclosures were reported.
Publication Details
Title: “Volatile Employment and Health Insurance Gaps Among Low- and Middle-Income Workers.” Authors: Vineeth Amba, MD, MPH; Renuka Tipirneni, MD, MSc; John Z. Ayanian, MD, MPP; Sumit Agarwal, MD, MPH, PhD. Published in JAMA Network Open, 2026;9(9):e2633883. DOI: 10.1001/jamanetworkopen.2026.33883. The article is open access under a CC-BY License.







