The core of my job is communication about what I call “panic headlines.” These are headlines — in newspapers, on Instagram — that provoke panic in parents. Often it’s along the lines of “New study says…” or “Experts claim…” followed by a terrifying claim (screens cause developmental delays, hot dogs linked with cancer).
In a very large number of cases, the panic these produce is misplaced and I spend a lot of time talking people down. A result of this is that people often ask me if there is anything I do worry about. The answer is yes: my two big safety worries with kids are cars and swimming pools. These are both sources of sizable risk for kids, but neither of them gets much panic-headline coverage, likely because they aren’t new and surprising.
Panic headlines instead tend to focus on things that you aren’t currently thinking about but (according to the headlines, anyway) should be. When processing new panic headlines, I end up pulling heavily on communication about one of two things: risk or uncertainty. In my experience, much of the panic comes from our difficulty in understanding small risks and from the challenge of living with uncertainty.
Today I want to flesh out both of those ideas, with some tools to discipline our thinking.
Understanding risk
People are extremely bad at understanding probabilities. This is especially true for very small or very large probabilities. Research in economics and psychology has shown us that people overestimate small probabilities and underestimate large ones.
Take the small-probability case. Let’s say I tell you there is a chance that something happens 1 in 1,000 times, and a chance some other event happens 1 in 5,000 times. These are actually quite different — one happens five times as often as the other — but for most of us, they feel similar. They are both small, but they aren’t zero. That “1”: it’s why we play the lottery! Even a chance of 1 in 80 million — you’re telling me there’s a chance! This internal psychology causes us to process even really unlikely events (like the lottery) as similarly likely to an event that happens, say, 1 in 1,000 times.
This confusion about small probabilities can push us to give attention to risks that may not be worth our time. Recently, in a Q&A and on Instagram, I talked about the risk of injury for children on airplanes. A comprehensive paper had estimated the risk of injury to a child under 2 at about 1 in 250,000; some of these injuries likely would have been prevented if the child was in a car seat in their own seat. So purchasing a seat could lower the injury risk by as much as 1 in 250,000 (assuming all injuries would have been prevented).
A gut reaction to this (one I got on Instagram!) is to say, Any risk reduction is worth it for your child. I can see the logic here, but the reality is that this risk is just extremely small. The annual risk of car accident injury for a child under 14 is about 1 in 500. The annual risk of a cancer diagnosis for a child under 5 is about 1 in 4,500.
Does this mean you shouldn’t use a car seat on an airplane? No — if you want to, go ahead, and there may be other reasons to do so. But making a good decision about tradeoffs here does require actually understanding the size of the risk.
How to conceptualize risk
Risk is not intuitive to people. Even for people who do this as their job, it’s not something our brains naturally “get.” Which means you need to actively find approaches that work for you to turn probabilities into something concrete that you can understand. The two approaches I like the best: other common comparisons, and conversion to time.
- Other comparisons: A simple way to think about risk is to ask what other events occur with this probability. It’s not an exact science, but articles like this one (admittedly, written for entertainment) can provide some context for probabilities. One in 250,000 is on the lower end of the estimated risk of being killed by a meteorite. A related form of this comparison is to take some risk that you do feel comfortable understanding, like the risk of car accidents, and use it as a benchmark.
- Time: This is my go-to: convert risks into time. Ask, If I took this risk every single day, how long would it take before I would expect the event to happen once? For something with a risk of 1 in 10, you’d expect it about every 10 days. For a risk of 1 in 100, it’s more like every 3 months; 1 in 1,000, every 3 years. For comparisons like 1 in 1,000 versus 1 in 5,000 — that’s every 3 years versus every 15 years. And those feel (at least to me) very different. Note: A daily risk of 1 in 250,000 translates to once every 684 years.
Bottom line on risk: A lot of panic ensues when we do not correctly conceptualize risks, especially when they are small. The first step is to discipline your risk thinking.
Living with uncertainty
Although we often use them interchangeably, risk and uncertainty are different. “Risk” means a defined chance of something happening or not — you may be uncertain about which outcome will occur, but you’re certain about the possible outcomes and their probabilities of happening. “Uncertainty” — at least formally, to an economist — means we do not even know the probabilities that we’re facing and, sometimes, we don’t know what the possible outcomes are. With uncertainty, we often cannot even get our heads around what we are facing, and that’s a bigger challenge than understanding probabilities.
A great example of a recent panic headline driven by uncertainty is “Costco sued for toxic chemicals in its baby wipes.” The story behind this is a class-action suit that claims too-high levels of the “forever chemicals” PFAS were found in a widely used Costco brand of baby wipes. This headline has uncertainty on many levels: it’s not clear to what degree we should be worried about PFAS in general (what the risks are, how large they are); it’s not clear what the levels are in the wipes; it’s not clear how much the wipes would actually transfer PFAS to a baby.
In the absence of these answers, we fill in with (often) worst-case scenarios. It’s easy to think, “This must be serious — it’s a lawsuit” (though worth remembering that there are a lot of reasons to file class-action lawsuits, and these same filers have other ongoing lawsuits against Costco…). Precisely because there is so much uncertainty, we cannot put these fears to rest with facts or logic. Thinking more about it will not help.
A similar case can be made for many of the headlines we see about screen time for little kids. You read something like “Screen time linked to developmental delays,” and there are a million questions: what kind of delays (there is no clear way to translate many of the metrics in these papers into real-world outcomes); for which children (most of these studies have heterogeneity across kids); what kind of screens (almost never measured). Then people like me come along and tell you, hey, actually, those studies just show correlation, not causation. It’s entirely possible that there is literally no effect at all on anyone. But … we can’t know for sure, because the study isn’t able to show us. It’s uncertainty piled on uncertainty.
Unlike with risk, there aren’t good tools for unpacking uncertainty — as noted, thinking more will not help. What we may have are tools for disciplining our interactions with these problems.
A first tool: Remember that in nearly all of these cases — all of the ones that come to mind, like ultra-processed food, microplastics, PFAS, screen time — the risks we are talking about are definitely small. If screen time caused enormous developmental delays in most children or even a small share, we would likely know it. “Small” isn’t a number, but I do think it’s useful to keep in perspective that a bag of Doritos, or an applesauce pouch, or a brand of wipes, or even an hour a day of television is extremely unlikely to matter much in the grand scheme of things.
A second tool: Ask yourself, What, if anything, can I do about this, and how hard would it be to do it? Sometimes there is an easy change that can remove the fears. If there is another, equal-priced brand of wipes that you’re indifferent about, you may as well switch. Often, though, there really isn’t anything realistic you can do. Think about panic headlines about, say, microplastics in the water supply. Moving away from the municipal water supply isn’t a feasible option for most of us.
We can advocate (and vote) for more water supply regulations, but at the end of the day, we have to realize that right now this is a risk we have to live with. I understand that this may sound callous and defeatist, but spiraling anxiety doesn’t actually help fix any problems.
The hardest cases are ones where there is something you could change but it would make other things worse. You could quit screen time cold turkey, but it would make your life much harder and mean more frustration with your kids. You could switch wipes, but it would cost more and your kid would like them less, making diaper changes less fun and eating into funds that could be used for other things. I would offer you the mantra: People face tradeoffs. No option is perfect or without downsides. You have to pick one and move forward.
Ultimately, this is about living with uncertainty. You will almost certainly never know whether you made the right choice in any of these many, many situations. Often, all we can do is embrace the uncertainty and try to move on.
Community Guidelines
Log in
This is all very good advice on understanding risk and uncertainty! One small technicality that I think is somewhat important: The number of days for a “1 in [frequency] days” thing to occur at least once is, in some sense, never, since there’s always a chance of it not occurring on each trial. So that leads to a slight recalculation of your heuristic of using the frequency: The number of days until there’s a greater than 50-50 chance of a “1 in [frequency]” event happening is usually 63% of that frequency. The larger the [frequency], the more accurate that rule of thumb will be.
Re: “if this were a big risk we would know it already”, how do we know that we aren’t already seeing effects of some of the stuff like microplastics/PFAS/ultraprocessed foods in the rising cancer rates among young people, etc? We don’t know why some of these trends are happening, so dismissively saying we can already guess the risks are small seems like even too much certainty.