The research, published in SSM – Population Health, was conducted by Joachim Gerich of the University of Linz and colleagues at the Federal Institute for Occupational Safety and Health in Germany. Rather than asking workers retrospectively how many days they had dragged themselves into the office while unwell, a method prone to memory distortion, the team deployed a factorial survey experiment. More than 3,500 employees, drawn as a follow-up to the nationally representative BIBB/BAuA Employment Survey 2024, each evaluated ten hypothetical scenarios in which they woke up feeling unwell. The scenarios systematically varied symptoms, body temperature, infectiousness, workload, team dynamics, attendance pressure, weekday and even weather, and respondents chose between calling in sick, going to work on-site, or working remotely if their job allowed it.

The results were striking in their clarity. Body temperature emerged as the dominant predictor of attendance behavior, accounting for by far the largest share of explained variance. At a mildly elevated 37.5 degrees Celsius, the probability of attending work on-site dropped by roughly 15 percentage points compared with normal temperature. At 38.3 degrees Celsius, a genuine fever, the probability collapsed by 54 percentage points. Strong cold symptoms, such as feeling weak and exhausted with a severe headache, reduced the likelihood of working by about 12 percentage points. No other situational factor came close to this explanatory power, confirming the classical model of Steers and Rhodes from 1978, which posits that ill-health acts as the central gatekeeper of attendance decisions.

But the study’s most consequential finding concerns what happens when the feverish employee has a laptop and a home office. Among respondents with a remote work option, a fever of 38.3 degrees still pushed 66 percent toward sick leave, but 28 percent chose to work from home and a stubborn 7 percent even went to the physical workplace. At the milder temperature of 37.5 degrees, remote workers were actually more likely to work from home than to call in sick. This pattern supports the theoretical argument of Fiorini that the availability of remote work raises the threshold of illness severity at which employees perceive themselves as sick enough to rest. In effect, the home office has become a middle path that absorbs many of the mild and moderate illness episodes that would once have ended in either full attendance or full absence.

The experiment also illuminated how contextual pressures operate only within a specific window of ambiguity. Attendance pressure from supervisors, the fear of overburdening colleagues, team cohesion and infectiousness all significantly shaped decisions, but their effects were concentrated at normal or moderately elevated temperatures, where the right course of action is genuinely unclear. When symptoms were very mild or very severe, these factors faded into irrelevance, echoing a hierarchical model of decision-making in which people skip deliberative stages when the answer seems obvious. Notably, high attendance pressure increased on-site presenteeism but left virtual presenteeism untouched, suggesting that managerial pressure operates primarily by pulling bodies into buildings rather than by extracting work in general.

Social dynamics within teams proved more influential than individual workload. When respondents imagined that their absence would significantly increase their colleagues’ workload, they became markedly more likely to work while ill, both on-site and remotely. Conversely, when they faced working alongside colleagues they did not get along with, the probability of calling in sick rose. Interestingly, the much-discussed Monday effect, the tendency for absence rates to spike at the start of the week in observational data, vanished entirely in the experimental setting, lending weight to the argument that Monday absences largely reflect illnesses that begin over the weekend rather than motivational shirking.

Perhaps the most sobering result is how persistent presenteeism remains even under objectively severe conditions. Across the full sample, the overall probability of working in some form, whether on-site or remotely, still reached 30 percent at a body temperature of 38.3 degrees Celsius. Men were slightly more prone to this than women, showing 32 percent presenteeism at high fever compared with 28 percent for women, and men with remote options favored virtual presenteeism while women more often chose the on-site route. Given that standard medical advice for acute febrile illness is rest and recovery, the authors caution that this persistence may deliver short-term productivity at the cost of long-term health depletion, a trade-off documented in prior research linking working while ill to serious cardiovascular events.

The motivational analysis added a revealing psychological layer. Respondents who chose to work despite illness rated their reasons on a scale, and the pattern shifted with severity. Inner convictions and principles were rated as the most important motive overall, followed by concern for the well-being of others and the accumulation of unfinished work. Crucially, as body temperature rose, those who still worked attached escalating importance to self-oriented and avoidance-oriented motives: protecting their reputation, advancing their career, avoiding unpleasant conversations, and preventing work from piling up. Prosocial concern for others, by contrast, remained flat across temperature levels. In other words, the sicker people were, the more their decision to work reflected personal disposition and identity rather than genuine consideration for colleagues or customers.

The study’s design carries both strengths and limits. The factorial survey approach, with its D-efficient orthogonal design and random assignment of scenarios, allows causal inference about situational effects that retrospective surveys cannot provide, and the preregistered analysis strengthens confidence in the findings. Yet hypothetical vignettes measure stated intentions rather than actual behavior, and the scenarios focused on acute cold-like illness, so the conclusions may not extend to chronic conditions, mental health problems or episodic diseases where working through symptoms can sometimes be adaptive. Factors such as organizational culture, leadership style and caregiving responsibilities also lay outside the experimental frame.

The implications for occupational and public health are nonetheless significant. Remote work appears to reduce outright absence in some situations, but it also seems to normalize working through mild illness and to blur the boundary between recovery and labor. As flexible arrangements become a permanent fixture of working life, organizations may need to make explicit that a moderately elevated temperature is a signal to rest, not an invitation to log in from the sofa. The German data suggest that without such norms, the flexibility that protects teams from infection and eases the daily commute may simultaneously erode the recovery time on which long-term health and productivity ultimately depend.

Subject of Research: Situational factors influencing sickness presenteeism and virtual versus on-site work attendance decisions among German employees

Article Title: The situational contingency of virtual and on-site workplace attendance decisions in Germany: a vignette-based approach

Article References: Gerich, J., Herdegen, N., Koch, A., Kroczek, M., Kugler, P., Meyer, S.-C., Perleberg, M., & Späth, J. (2026). The situational contingency of virtual and on-site workplace attendance decisions in Germany: a vignette-based approach. SSM – Population Health, Article 101979. https://doi.org/10.1016/j.ssmph.2026.101979

Image Credits: AI Generated

DOI: 10.1016/j.ssmph.2026.101979

Keywords: sickness presenteeism, remote work, attendance decisions, factorial survey, body temperature, occupational health, Germany, workplace behavior, virtual presenteeism, sickness absence, team cohesion, attendance pressure