Every now and then, in a conversation about health, I run across someone who will offer up, “My granny smoked a pack of cigarettes a day her whole life and she lived to be 90!” I’m not sure if it it’s meant as an endorsement of smoking or a smackdown of the conventional wisdom on healthy living, but it’s always a line delivered with passion.

Admittedly, conventional wisdom, even in the medical establishment, has gotten a lot wrong. Many of us living today remember the low fat diet craze, which sparked an epidemic of obesity and diabetes. Lobotomies were once seen as an acceptable solution for people experiencing phobias, drunkenness, homosexuality, or women with mood swings. The inventor, if you can call randomly sticking sharp objects into the brain an invention, was actually awarded a Nobel Prize in 1949.​

So yeah, there’s a lot the medical establishment has gotten wrong, and I didn’t even get to the leeches! (A future newsletter, I promise.)

When trust in our leaders and institutions falters, rightly or wrongly, we tend to rely on personal experience to gauge risk instead. We are overwhelming social decision makers. If you have a chain smoking granny who lives a long time, you might conclude smoking isn’t as harmful as the experts make it out to be. But if your mother smoked, as mine did, and went on to both get lung cancer and die from it at a young age, you might come to a very different conclusion.

But this social strategy of risk assessment is itself risky, particularly in situations where there is a delay in observing the outcome and its severity.

The data tells us that only one out of every five smokers ultimately develops lung cancer. But lung cancer, at present, is a virtual death sentence: 85% die within five years of their diagnosis.

Smoking has other adverse outcomes beyond lung cancer, but for the purposes of this example, let’s keep it simple. Roughly 17 out of every 100 smokers will die of lung cancer.

Maybe those sound like pretty good odds. They aren’t.

My personal observation is that most people have a mental rubric where anything under 10% is rare, anything over 90% is likely, and everything in between is either unlikely or somewhat likely.

That is, how do I put this politely, really skewed. In medicine, the likelihood of an adverse outcome is described this way.

Based on this table, we might categorize a 17% chance of dying from lung cancer as a likely outcome. But likelihood isn’t the only factor we need to consider. When we also consider the consequences of a particular outcome, as shown in this standard risk matrix, we see that smoking is extremely risky (assuming you agree that death is a catastrophic outcome).

My point is, even with the data, most people underestimate risk by a lot.

If I told you, “Did you know that four out of every five smokers will never get lung cancer?” Chances are you’d say, “Wow, I thought smoking was a lot more dangerous than that.” (Note that I simply flipped the data I already gave you—if one in five develop cancer, that means four out of five don’t.)

So whether we’re observing our peers or looking at the data, we struggle to accurately understand risk because we have a skewed view of likelihood and tend to ignore the severity of the consequences all together. It gets even harder when there is a significant gap between behavior and outcome, like with smoking and lung cancer. Imagine the proverbial “If all your friends were jumping off a bridge” problem, but you didn’t find out what happened to them for twenty years. Chances are, you'd jump too.

Being aware of this limitation, while sobering, is helpful. I like to remind myself that our inability to effectively gauge risk is a human frailty, not an individual one, so we can give the medical establishment, business leaders, and everyone else who gets it wrong a bit of a break. The fact is we can neither wholly trust our leaders nor ourselves when it comes to risk.

Which leads us, I believe, to two important behavioral shifts:

1) We shift our focus from our perception of likelihood to consequence. As you can see from the risk matrix, when the consequence is major or catastrophic, even unlikely events are highly risky.

2) When there is the potential for major consequences, we should err towards the more cautious approach or decision. In these situations, it’s especially important not to rely on what everyone else is doing to guide your behavior. Assume everyone else likely doesn't see the risk.

For example, someone might question whether there’s any point in changing their behavior in relation to climate change, because it’s unlikely that one person can have an impact on global warming. This is the equivalent of focusing on likelihood. When we shift our focus to the consequences of climate change however, the cautionary approach tell us it’s less risky to change our behavior in a way that reduces our carbon footprint than to continue the status quo.

I won’t sugarcoat it, this can be hard. One, because we tend to justify whatever decision we want to make anyway. And two, because we are social decision makers, we overemphasize social risks (What if people think I’m weird or overreacting?) over other consequences (What if the planet warms beyond the 1.5 degree threshold in my lifetime?).

We don’t talk about it much, but being cautious actually requires quite a bit of courage.

But by definition, if you want to be a leader, you have to be ahead of the masses. When you find yourself at the bridge, you have to be willing to tell not just strangers but your friends, “Nope. This is too risky. I’m not going to jump.”

And when you walk away, you hope someone follows you. Because if they do, you’re a hero, even if they never call you that.

Have something you want to say about this post? Just reply to this email or join the conversation over on LinkedIn.