Fig. 2: Work hours relative to full-time employed.

Hours worked is defined in full-time equivalents (FTE), which implies that a value of one represents a full-time job, and zero that the individual does not work. Panel (a) People with HIV and matched controls shows the average hours worked over time relative to diagnosis (at t = 0) for people with HIV (solid blue line) and matched controls (dashed gray line). Panel (b) People with HIV and controls by disease stage makes an additional distinction between people with HIV and matched controls depending on stage at diagnosis: non-late (solid blue line and dashed light gray line, respectively) and late (solid orange line and dark gray dashed line, respectively). Panels (c) Difference-in-differences estimates and (d) Difference-in-differences estimates by stage report event studies estimated using the Callaway and Sant’Anna estimator⁵⁰. Each dot represents a point estimate for a specific period relative to diagnosis, and the bars represent the 95% confidence intervals, where t = 0 represents the year of diagnosis. In panel (c), N = 61,974 observations at period t-1. In panel (d), N = 33,792 observations at period t-1 for individuals not diagnosed late and N = 24,937 observations at period t-1 for individuals diagnosed late. The period t = −1 is the reference period and is standardized to zero. Data on hours worked is only available starting in 2006, which restricts the number of years before diagnosis that we can observe to four. Source data are provided as a Source Data file.