Covid arithmetic for anxious parents

By Bryan Caplan

Starting March 1, masks are optional for all K-12 students in Virginia public schools. That includes my daughter, who after a year and a half of homeschool decided to rejoin her friends in 4th grade. Since plenty of parents are terrified of this liberalization, please let me walk through the risk arithmetic, then put the numbers in perspective.

Photo by CDC

The question, to be blunt, is: What are the odds that not wearing a mask will lead a child to die of Covid?

1. Let’s start with the Infection Fatality Rate (IFR) of the original version of the disease. Estimates vary widely by country, but 6 in 1000 remains a reasonable overall estimate.

2. In the Omicron era, U.S. cases rose roughly ten-fold, but deaths only multiplied by 2.5x. This implies an IFR of roughly one-fourth of the original, down to 3 in 2000.

3. IFR roughly triples for every decades of age. If you’re 50, you’re near the population average. For a child of ten, you can divide the average risk by 3*3*3*3=81, bringing us to about 1 in 50,000.

4. Vaccines now seem about 80% effective. For vaccinated kids, the odds of death therefore fall to one 1 in 250,000.

5. Adults often speak as if properly-worn masks are 100% effective. Actual Randomized Controlled Trials (RCTs), however, find far lower efficacy:

In sum, of the 14 RCTs that have tested the effectiveness of masks in preventing the transmission of respiratory viruses, three suggest, but do not provide any statistically significant evidence in intention-to-treat analysis, that masks might be useful. The other eleven suggest that masks are either useless—whether compared with no masks or because they appear not to add to good hand hygiene alone—or actually counterproductive. 

Reading through the evidence, a 10% mask efficacy remains credible. Translation: There is a 1 in 2,500,000 chance that regular mask-wearing will save a vaccinated child’s life.

6. The hardest-to-resolve risk issue, strangely, is the odds that a child will even be exposed to Covid. Cases have fluctuated wildly – an Everest spike this holiday season, followed by a symmetric crash. Now cases are really low again, and still falling. Maybe we’ve finally hit herd immunity; perhaps we’ll see another spike, or a never-ending series of spikes. In light of this uncertainty, let’s just say there’s a base 50% chance that kids who go to school will still catch Covid one way or another, bringing the odds that wearing a mask will save a child’s life down to 1 in 5,000,000.

7. By way of comparison, a child’s annual risk of death in an auto accident is about 1 in 50,000. That’s 100 times higher than the non-masking risk. Honestly, if you’re worried about your kids’ safety, it would make vastly more sense to just drive less. Most obviously, you could withdraw your kids from sports, scouting, dance, and travel.

8. If you protest, “That’s a crazy overreaction; I want my kid to actually enjoy his life, not just stay alive,” I agree. And that is a great argument against masking as well. After all, masks aren’t merely inconvenient. They are dehumanizing. There is a reason even mask hard-liners unmask around immediate family members. Namely: Covering your face emotionally cuts you off from other human beings. If masking reduced kids’ risk of death by 1 in 100, you could reasonably respond, “Tragically, this dehumanization is worth the price.” But not if masking reduces the risk by 1 in 100,000, much less 1 in 5,000,000.

9. I’ve focused on death, but what about other Covid-related harms? Key point: Almost everything that kills can also maim. Car accidents are once again the go-to example. We have much better data on deaths, but car crashes also cause paralysis, loss of limbs, brain damage, and more. If that isn’t enough to make you sharply cut back on driving, fear of long Covid shouldn’t sway you much, either. If anything, you should be more worried about non-death dangers of driving, because a large share of “long Covid” symptoms seem to be psychosomatic. In the age of Covid, we tend to blame mysterious health troubles on Covid.

10. Does this mean that schools have been hysterically overreacting from the get-go?! My answer: Probably. The original strain was four times deadlier, and vaccines are about 80% effective, so the early risk was 20x higher than today. But for kids, the benefit of masks was comparable to taking 20% off their risk of dying in a car accident. The health gain of strict isolation was probably several times bigger, but strict isolation is terrible. Remember: Solitary confinement is a form of torture. While isolating with your family is hardly solitary confinement, it’s in the same ballpark. Especially for only children!

11. What would the ideal school response to Covid have been back in March, 2020? My considered judgment: Wait two weeks to double-check the numbers. Advise very high-risk kids to stay home. Have very high-risk teachers run zoom classrooms for high-risk kids. Otherwise, re-open without abnormal precautions.

12. Can you quibble with my numbers? Of course; they’re all estimates. But no credible modification will change the fact that risks of Covid to kids have gone from tolerable to microscopic.

13. Should we really make decisions about children’s health using math? Well, why do you think we teach kids math? The main point of math is to improve human decision-making. Life entails risk. No one is perfectly safe. Not you, not me, not your kids, not my kids. If you can moderately increase safety with mild precautions – like seatbelts – great. If you can drastically increase safety with strong precautions, maybe that’s the right path, too. But microscopically increasing safety with strong precautions isn’t just a bad idea. It is childish.

14. Clever readers will object, “We’re not masking kids to protect kids. We’re masking kids to protect adults. The adults who work in schools. The adults kids live with.” I say this is greatly overstated. Yes, many adults worry about catching Covid from kids. But actions speak louder than words. And what most adults’ actions say is that they worry more about kids’ health than their own. Plenty of parents will go to an unmasked party, but still insist the Girl Scouts mask at their meetings. Much the same goes for school employees. The fundamental problem is that adults are “thinking” emotionally about issues that you can only understand mathematically.

15. Yes, the risk to adults is much higher than the risk to kids; as I said, every ten years of age roughly triples the danger. But when you multiply a microscopic risk a hundred-fold, you still end up with a minor risk. Yes, you could accept a dehumanized existence to minimize your disease risk, but you would be foolish to do so.

16. The most “adult” thing for parents and teachers to do on this first week of March is walk kids through the Covid math. Even third-graders should be able to follow it. And if you really want to show your maturity, you should confess that for the last two years, most adults have been acting like children. Life gave us a math project, yet we acted like it was a poetry assignment. Our kids deserve better. Let’s start giving it to them!

P.S. Thanks to Charley Hooper for pointing out a few mistakes.


Bryan Caplan is a Professor of Economics at George Mason University, and the New York Times Bestselling author of The Myth of the Rational Voter, Selfish Reasons to Have More Kids, The Case Against Education, Open Borders, and Labor Econ Versus the World.  He is also a Visiting Senior Scholar at the Salem Center for Policy at the University of Texas, Affiliated Faculty at the Political Economy Research Institute at Middle Tennessee State University, Senior Research Fellow at the Mercatus Center, and Adjunct Scholar at the Cato Institute.  He blogged for EconLog from 2005-2022.