On paid sick days

2/10/2016 09:13:00 AM
The United States is one of the few developed countries that still does not have a sick leave program. Most countries have a publicly funded sickness program that pays workers not to go to work. Over 102 countries offer workers more than a month of sick leave, at varying levels of wage replacement. In the US, by contrast, sick leave programs are administered by employers, if they choose to have one at all, and typically only offer 5 days off.

Thus, there's a substantial movement in the US to make sick leave benefits universal by mandating that all employers provide at least 5 paid sick days off to all employees. There are lots of benefits to having sick days. For one, it's the humanitarian thing to do—when someone is violently ill, we should let them stay home and take care of themselves, or to go to the doctor if necessary. For another, having a sick person at work can, for example, increase risk of workplace accidents. There are lots of other benefits as well, but one that is most often touted is that letting sick peoples stay home will reduce the spread of diseases in the workplace.

Unfortunately, I think that last particular benefit is overstated—US-style sick leave policies have only minimal effect on the incidence of disease.

The problem is that the typical sick leave policy allows only 5 days off, and even then we're faced with strong social norms prohibiting workers from using them effectively to prevent contagion. Your employer and co-workers will suspect you of shirking if you call in sick but don't sound violently ill, and you're expected to only take one sick day—2 at the very most—for a particular illness.

Surveys show that for the flu, one of the more aggressive and dangerous of the common illnesses, workers with access to paid sick days take fewer than 1.8 days off on average. The problem is that the flu remains contagious for 5 days, and up to 2 weeks for some people, so taking only 1 or 2 days off from work does not substantially reduce your co-workers' exposure. Even taking all 5 days would not eliminate the risk of spreading flu at work. The result is that access to paid sick days only reduces workplace transmission by less than 6 percent.

Aside from food poisoning and hospital-acquired infections, most common contagious illnesses these days are viral and not amenable to treatment with antibiotics. Most of these viral illnesses have a pretty similar menu of symptoms, differing in severity, including fever, congestion, runny nose, sore throat, nausea, and diarrhea. But epidemiologically, they're all over the map. Some viruses become contagious before people develop any symptoms, others remain contagious after the symptoms disappear, and many viruses can be transmitted by people who never develop any symptoms. The notorious norovirus that infected Chipotle restaurants on two recent occasions, for example, remains contagious 3 days to 2 weeks after their symptoms go away, and 30 percent of people who have norovirus infections never develop any symptoms at all. Most of these viruses remain contagious for longer than the 1 or 2 days workers typically take off from work.

Having more sick days could help. In other countries with public sick leave programs, workers typically take 5 to 10 days off a year for sickness, much more than the 1.8 in the US. And some theoretical and empirical estimates in the US suggest that having access to 7 to 10 days off a year could be associated with a much larger reduction in workplace transmission, especially during larger epidemics. But even then, sickness leave policies will leave large gaps as workers come to work whenever they lack symptoms, even if they're still contagious.

Why is this so hard?

I think it helps to make a comparison to vaccination programs. Vaccines obviously aim to give individual-level immunity to the individuals who get it, but when administered to the entire population they also provide a second level of protection known as "herd immunity," where the transmission of the virus is impeded so much that it is unable to infect even those without individual-level immunities. Even with vaccines that are highly effective at producing individual-level immunity, you often still have to vaccinate more than 90 percent of the population before you start to see significant herd immunity effects. Paid sick leave policies aim to create a herd immunity effect—impede the transmission of illnesses to the non-immune—without any of the individual-level immunity that makes this work for vaccines. If we really want to see a significant effect here, we'd need to dramatically ramp up the coverage of paid sick days. Not merely extending those 5 days off to the whole population, but also increasing the number of days and ensuring that workers take weeks, not days, off whenever they get sick. Without that dramatic step, I doubt most work places will see much effect on illnesses from their sick leave policies.

Update 2-24-2016: post has been updated to add more links to sources.