The Price of Staying Home

A paid sick leave rule does not simply state whether a worker is covered; it sets the price of staying home at different points in an illness. This chapter focuses on waiting periods, medical monitoring, replacement rates, and benefit duration because those design choices determine whether the system deters unnecessary absence or makes genuine and contagious illness harder to leave. The useful judgment is directional: before asking whether a rule is generous, ask which behavior its timing and conditions make more attractive.

Rules meant to prevent abuse can raise the cost of genuine absence

The simplest anti-moral-hazard device is a waiting period. Benefits begin only after one or several days, so a worker who takes a very short absence bears part or all of the initial cost. The intended mechanism is clear: if some people would claim leave whenever it is convenient, an unpaid first day may discourage those claims.

But the same rule applies to a worker who is genuinely ill. If the first day is unpaid, staying home becomes more expensive precisely when the worker is deciding whether to report symptoms, rest, or attend work. A person with a contagious condition may decide that the lost income is more serious than the risk of exposing others. The waiting period can therefore reduce one kind of absence while increasing the chance of work while sick. It does not distinguish perfectly between a shirker and a person whose illness is real.

The IZA review describes a Swedish randomized field experiment in which a one-day waiting period was abolished and replacement rates for longer durations were increased. The combined change reduced sick-leave rates overall, and employees on leave returned to work more quickly after the waiting day disappeared. The interpretation is not that more generous benefits always reduce absence. Rather, the old waiting day may have created a discontinuity: once a worker had paid the price of the first day, they had an incentive to prolong an existing spell so that a relapse would not require paying for another uncompensated day.

That result makes the design problem more subtle than “waiting periods reduce claims.” A waiting day can deter some short, unnecessary absences, but it can also alter the length and timing of spells. It may push a worker to attend on the first day of an illness, return before recovery, or remain away longer after the initial cost has already been paid. The Swedish evidence is useful because it demonstrates that removing a restriction can change both the amount of absence and the way workers move through a spell. It does not isolate the waiting-day effect by itself: the experiment also changed replacement rates.

Medical monitoring creates a related trade-off. A doctor’s note or another certification requirement is meant to verify that absence reflects limited work capacity rather than preference. More restrictive monitoring can reduce absence and the number of workers who shirk. But verification is not free or perfectly accurate. It imposes time and financial costs on the worker, uses medical resources, and requires someone to judge a condition that may be difficult to observe from a short consultation.

The IZA review emphasizes the behavioral consequence of stricter monitoring: it can increase the share of workers who come to work sick, spread disease, work at low productivity, or experience work-related accidents. If obtaining documentation takes time, money, or access to a clinician, the worker faces an additional hurdle before receiving protection.