Risk is hard to understand.

For example, if I said that we should be more worried right now about dying from a slip, trip or stumble on level ground than about dying from swine flu, you might think I’m nuts. But I’m not. Statistically, at this moment, we’re about 700 times more likely to die stumbling than from getting swine flu. If swine flu fatalities were a quart of oil, then stumbling fatalities would be about 3 oil drums.

Even if it’s true that tripping is more lethal than swine flu, it’s hard to fathom. H1N1 just seems much scarier right now. Why is that?

Mainly, it’s our brains. They’re not wired to be objective. The analytical response to risk–the part when we come to understand how likely or unlikely something is to occur–is subsequent and sometimes subservient to our emotional response to risk–the part when we’re scared out of our wits.

That is, first we fear, then we figure. Evolutionarily speaking, this makes sense. For the species to survive, we had to meet every uncertainty with caution.

Unfortunately, we didn’t evolve in the presence of mass media, and mass media really messes with our brains. Images, sounds, access to rare but spectacular events–and the ability to instantly share all of that with each other–actually reinforces our emotional response to risk. It makes it harder for us to analyze the risk objectively, to put it in a realistic perspective.

This is the reason our hearts pound like a timpani when we get on a roller coaster and not when we slip into a bath tub. You’ve probably heard about decapitations and deaths on roller coasters. You haven’t heard about drowning in bathtubs. And yet, while about one in 150 million riders die on roller coasters, the chances of drowning in a bathtub are just under one in a million, making it more than a hundred times more common.

(It should be noted that all of the risks we talk about here are
extremely rare; none is likely to happen to you in your lifetime. In
fact, in your lifetime, you have a one in three chance of accidental
death. Or put another way, you have a two in three chance of not dying by accident).

Likewise, when we hear of a child dying after contracting swine flu, it’s tragic and sad and no amount of statistical analysis can convince us that, while that’s heart wrenching, it’s also a freakishly rare event. This is a problem risk experts call “imaging the numerator.” The threat seems worse because we can imagine a dying person. Even if it’s a one-in-a-trillion, we see the numerator, the one, because we can relate to that. It’s much harder to relate to the fact that one swine flu death represents three millionths of a percent of the US population.

In a non-mass media world, the vast majority (nearly 100 percent) of the population would never know it happened. But now, for the first time in history, we have unfettered access to a seemingly infinite amount of stimuli that scare the bejeezus out of us, that help us image the numerator and, in turn, make it harder for us to logically analyze the real risk facing us.

All of which would be only mildly interesting if it ended there, but of course it doesn’t. The big problem with being guided by our emotional response to risk is we make decisions–parenting, educational, policy, strategic business decisions–based on it.

For example, parents take kids off of buses and drive them to school because they hear about, read about, see news reports on or even witness bus accidents in which children die.

It’s not hard to understand why parents would do that. And yet, it’s the wrong decision. Buses are seven times as safe as cars, and one of the safest forms of road travel going.

This flu threat could create similar counter-intuitive decisions, but on a larger scale and with rippling effects. Policy is created in the emotional response to the flu. The flu interrupts the business of the planet and forces reallocation of resources toward the new threat. But are all those resources commensurate with the actual risk? 

If it turns out that the unending, often sensationalistic coverage of this latest flu virus was overstated–and it appears right now it was–that means decisions were made–closing schools for two weeks in Texas, shutting down business in Mexico–that were unproductive or counterproductive.

What’s more, it means resources were taken away from other threats in order to address this one.

For example, there have been zero deaths from swine flu in the workplace in the United States since the outbreak was first reported, about ten days ago. In that same ten days, there will be, on average, more than 20 workplace homicides. Workplace violence is, in fact, a far more serious risk to business than the flu.

It’s easy to be inefficient with resources. But if resources are, by definition, limited, then we have to be careful about making bad decisions with them in that moment we’re emotionally responding to risk. Before we take the kids off the bus, let’s see if that’s necessary, or even desirable.

There’s no magic formula to make allocation of resources perfectly commensurate with the risks they’re meant to offset. What’s more, there’s always the chance that some bit of analysis demonstrates that the risk in question is extremely rare, and yet it comes to pass anyway. That doesn’t make the analysis wrong, but it sure feels wrong when you’re the one who gets the swine flu.

To be clear, these risks need to be disclosed. Swine flu could get worse and it’s better to be prepared than not. The trick is how do we disclose responsibly, in a way that puts the risk in perspective? In a mass media world, that’s a quixotic challenge.

One of the best approaches to rare but spectacular risks like swine flu–for people and for businesses–is common sense. So wash your hands and tell your employees to wash theirs; if you’re sick, stay home. Discourage, or ban, presenteeism.

And try to keep things in perspective. Be aware but not afraid. The risk of swine flu is there, but it’s relatively small. Tiny, even.

(Many of the statistics used in this article came from the National Safety Council’s Odds of Dying chart for 2005, the most recent year data was available).