Volatile hours are a health exposure, and the bill arrives differently for men and women

Researchers at the University of Illinois Urbana-Champaign report that week-to-week swings in work hours, not only the total number of hours, track with worsening self-rated health among more than 61,000 US workers aged 18 to 64 ^[1][3]. The operationally important part is the split: according to the authors, men exposed to greater fluctuation were more likely to report deteriorating health and stay in their jobs, while women were more likely to reduce hours, suspend work or leave employment entirely for health reasons ^[4].

The paper, in SSM - Population Health, is by doctoral student Sinyee Qianyi Lu and sociology professor Tim Liao ^[1][2]. They pulled Current Population Survey basic monthly files and the CPS Annual Social and Economic Supplements from the University of Minnesota's IPUMS, linking individuals across the two datasets so that hours and health could be followed over two consecutive calendar years ^[9]. The CPS is a monthly household survey run jointly by the Census Bureau and the Bureau of Labor Statistics ^[3]. Health status and month-to-month hours were tracked over a 16-month span ^[5], built from two four-month observation windows two years apart with an eight-month unobserved gap between them, covering 2016-2018 and 2020-2024, so prepandemic, midpandemic and postpandemic years all sit in the pooled sample ^[10]. Health is self-rated, excellent to poor, coded into four outcomes: improved, stable, deteriorated, or health-related work limitation ^[11].

The sample construction is the argument. A baseline model contained 55,062 continuously employed people; the expanded model reintroduced 6,069 who reported a health-related work limitation in their second year ^[6] - working part time for health reasons, absent because of illness or medical problems, or out of the workforce entirely due to illness or disability ^[8]. That is roughly one in ten of the 61,131 total ^[7], and it is exactly the group prior studies often excluded ^[16]. As Lu puts it, treating health-related work limitation as an outcome rather than as sample attrition let the analysis follow people even when deteriorating health pushed them to cut back or exit, and expanding beyond the continuously employed "survivors" brings the wider at-risk population into view ^[14][15].

Among men, the researchers report that greater volatility in hours was associated with a gradual decline in self-rated health, most visibly among those working longer weekly hours, while they stayed employed ^[12]. The result as summarised is directional rather than quantified. Lu's framing is that job quality is not just pay, benefits or total hours but schedule stability and the health outcomes that follow, an area she describes as underexplored ^[13]. The study joins a growing literature on the health consequences of unstable work time ^[17].

Two things to watch. First, whether the volatility signal survives separation from pandemic-era disruption, given that 2020-2024 is pooled with 2016-2018 ^[10]. Second, whether the gendered exit pattern ^[4] shows up in employers' own data, since a firm that measures only headcount and average hours will see the women leaving and never see the variance that preceded it, nor the men accumulating health decline on the payroll.