Editor’s note: Burney Waring, the president of the local nonprofit Innovate Chatham, is a global consulting engineer who studies trends and data. He shared this look at COVID-19 risks for children with the News + Record.
Risks are especially difficult for humans to understand for multiple reasons. Let’s discuss the challenges, and use some techniques to try to make risk more understandable.
Specifically, let’s look at the risk of children not wearing masks in school. The risk will vary for each child, of course. Here we will strictly look at averages, and we will not make a recommendation about masking. We will simply give different ways to consider risk in decision-making.
Daniel Kahneman is a well-known psychologist, and economist who writes accessible, respected books about the psychology of risk. Kahneman finds that we tend to make almost all decisions with our fast gut instinct, and few with our slow thinking mind. It’s a good thing too — we could not survive if we had to stop and think carefully through every decision, every day. But obviously, we should try to do our best thinking about our most important decisions.
Psychologists know that humans have natural, common flaws in how they think, and they call this “bias.” For example, we tend to exaggerate the risks of unfamiliar, or rare events (for example, shark attacks) and understate common risks (for example, heart disease) to which we have not personally been exposed.
We also have bias about very small and very large numbers. We evolved to understand and use the sorts of numbers that are used in counting and for dividing sets of items into a small numbers of parts. On the other hand, fractions in the millionths and numbers in the millions are far beyond our gut instincts.
Making decisions about COVID-19 is a perfect storm of bias: unfamiliar, with both tiny numbers and huge numbers, and with limited past experiences to train our individual gut instincts.
Some numbers: 970 children have died with COVID-19 to-date, according to the Centers for Disease Control and Prevention (CDC), out of a population of 73 million children in the U.S. That means that the average rate of children dying of COVID was 13 out of every million in the U.S., over the past two years. (Source: Provisional COVID-19 Deaths: Focus on Ages 0-18 Years; Data: Centers for Disease Control and Prevention; cdc.gov.)
While our gut instincts are lacking for tiny numbers like 13 in a million, humans do well visualizing fractions. Imagine an image showing the fraction of total U.S. deaths of children with COVID-19 to the total U.S. population of children with a tiny white square in the upper left-hand corner of an otherwise huge black rectangle representing the fraction of 13 out of a million. The simple average rate for 1 year would be half that, or 6.5 deaths per million, which will be useful for comparison.
Those are measured rates. From rates, we can estimate risks. If the same situation exists, the past rate and the future risk are expected to be the same. Because the overall measured death rate is dropping, and children are getting vaccinated, it is likely that the current risk is smaller and will continue to shrink in the future (unless some more dangerous variant arrives).
Another powerful bias: we tend to make decisions not on the risk itself, but by comparing risks. Cutting the risk of shark-bite deaths in half sounds great to our gut instinct. “Half the risk” is a statement of relative risk. But cutting shark-bite deaths in half would be perfectly trivial relative to cutting the risk of heart disease deaths in half, because the average absolute risk of death in the U.S. from a shark bite per year is about 1 in 300 million, while the average absolute risk of death in the U.S. from heart disease per year is 2,000 in a million (1 in 500).
(Source: Yearly Worldwide Shark Attack Summary – International Shark Attack File; ufl.edu; Heart Disease Facts, cdc.gov.)
Absolute risk is the only risk that can hurt you, and it deserves consideration in all decisions. If you decide the absolute risk is not important, then you stop worrying.
If you decide the absolute risk is important, only then should you consider relative risks. Our tendency to focus on comparing risks is a very strong bias. Paraphrasing Kahneman, our thinking brain is lazy whereas our gut feel is always ready and eager. So we probably tend to use relative risk because we have an instantly-available gut instinct for “half as much,” and find it necessary to think carefully about “1 in 300 million.”
The CDC recently found that mask wearing indoors reduced the rate of a positive test by 56%. That might mean that removing masks would increase the absolute risk to children from 6.5 to 15 deaths per million, each year. There are about 8,800 children in Chatham County Schools. So, if there is no mask-wearing in school, we would expect the deaths of children in CCS to increase from 0.057 children per year to 0.130 children per year. A different way to express this is that the risk would increase from an average of 1 death every 17 years to 1 death every 8 years.
(Source: Effectiveness of Face Mask or Respirator Use in Indoor Public Settings for Prevention of SARS-CoV-2 Infection — California, February–December 2021, MMWR; cdc.gov. Search for Public School Districts - District Detail for Chatham County Schools; ed.gov.)
If you consider the absolute risk described above to be important, then it may help when making decisions to consider the relative risk of other, more familiar dangers — for example riding in a vehicle. In 2020, 38,680 people died from vehicular accidents in the U.S., while we drove 2.8 trillion miles. So, the average rate of death for people who rode 10,000 miles was 137 in a million. From this we could conclude that car-riding 10,000 miles has around 20 times the average annual risk of death to a child from COVID-19. If so, it would be sensible to be much more worried about the danger of a child frequently riding in a vehicle than a child catching COVID-19.
A final risk comparison: Compare the rate of COVID deaths of children to those over age 50: about 1 in 140 people over the age of 50 have died over the course of the pandemic.
A final bias worth mentioning: When shown new information, humans tend to stick to their prior belief. That is called confirmation bias. If this analysis changed your mind about risks, that would be extraordinary.