Terror at 30,000 Feet

 | 

The old joke about the statistician who drowned in a lake with an average depth of one foot is a reminder that while the mathematics of probability theory are rock solid (er, within a certain range of error), the questions that the numbers attempt to illuminate are a bit more slippery. To put this in another way, a statistic is only as valid as the manner in which the question it tries to answer is framed. And there’s the rub: a question can be spun in such a way that the answer will confirm any sophistry.

This insight was recently brought home to me by Tyler Cowen’s wonderful Discover Your Inner Economist: Use Incentives to Fall in Love, Survive Your Next Meeting and Motivate Your Dentist. But even libertarian economists can fall prey to their inner biases. (I haven’t discovered whether Cowen calls himself a libertarian or not, but following Rush Limbaugh’s opinion that all economists worth their salt are libertarian, I suspect he is.)

At one point, Cowen briefly discusses fear of flying, citing various statistics that “prove” that flying is, hands down, much safer than driving a car. When one compares mortality rates per mile traveled and per passengers involved, the conventional figures decisively prove their point.

So why am I not scared of driving? As Ayn Rand famously stated, “Check your premises!”

Having taken flying lessons (and having had to land a single-engine plane that lost power), I have a slightly different take on the matter. A Cessna 150 with a perfectly centered dead engine practically lands itself, slowly gliding down at the proper angle, needing only a steady hand to keep it from diving into a stall. By comparison, a multi-engine jet with the reduced glide ratio that results from swept-back wings, and the out-of-balance weight and thrust from an off-center, suddenly faulty engine, almost requires a miracle to land safely.

Cowen, along with many others, believes that fear of flying is irrational. Now, I consider myself a rational empiricist, but when facing a flight, I gird my loins and make sure my affairs are in order. And I don’t think my fear is irrational. Yet I had never really tried to work out the problem until I read Tyler Cowen, who skewers popular fallacies as only a libertarian economist can. My conclusion is that he may have embraced a popular fallacy himself.

A stalled car engine is an inconvenience for, perhaps, half a dozen people at the most, while a stalled jet engine is a likely death sentence for hundreds of passengers. Having a pigeon fly into a car’s grille is startling, but it has far from the same consequences as having a pigeon fly into the cowling of a Rolls-Royce Trent 900 engine.

The questions I would pose to determine the safety of flying vs. driving would be: What percentage of mechanical malfunctions in cars result in fatalities? And how many fatalities? But what percentage in planes? I’m willing to bet that mechanical malfunctions (or operator errors) in an airplane cause way more fatalities than the same problems in a car. Different premise, different conclusion.




Share This

Comments

Andy Currie

As the last poster said, fear is not based on mathematical probabilities, it's an emotion and it's related to control over one's destiny. Passenger jet travel makes me uneasy for the same reason I'm a libertarian: I don't like surrendering all control over my life to a stranger.

I also have a pilot's license and when I am in the cockpit I am no less nervous than when driving a car. It is a more complex task, and it requires more concentration, perhaps, but I don't have to wonder what the guy up front is doing wrong because I am that guy.

It's also true that in the event of an airplane accident, the risk of death or serious injury is higher than with a car, but the risk of having that accident is lower. Much lower, in fact. But the knowledge that the consequences of a mishap are so high influences our intuitive fear for the same reason we get excited over a high lottery prize. The stakes are so high they can't be ignored.

Rail travel is statistically even safer than air. Even though rail accidents are more frequent, they are less likely to kill the passengers. But if you like the freedom to decide your own destiny, there is some uneasiness inherent in rail travel, too, since once you are aboard you have about as much control as you do in an airplane. Fortunately, Amtrak compensates for this by being more comfortable.

And you can always sit in your sleeper cabin and get drunk, too.

eeyn524

The difference is whether responsibility for your safety is partially in your own hands (driving) or entirely in someone else's (flying).

Being transported by a statistically average driver isn't one of the actual choices. Of course people might overestimate their own driving skills, but that's a different argument.

J Eyon

i'm glad you brought up this perspective - i'v often felt that the per mile comparison isn't the only way to judge things

reducing the study to per "mishap" is another way - whether the mishap is driver/pilot error or mechanical

or per "unit of time" - as opposed to "per unit of distance"

i'm sure there yet other numbers that mathematicians could crunch

i suspect most or all of these will still demonstrate that flying is safer - statistically - and explained by the greater care taken by average commercial airlines compared to the average driver - and to the more crowded conditions of driving - etc

but that won't trump the fear factor - which knows that the odds don't mean a thing - cuz in reality - a mishap will happen or not happen - 0% - or 100% - probabilities are simply a way of telling you whether to be surprised or not

Sevo

"I’m willing to bet that mechanical malfunctions (or operator errors) in an airplane cause way more fatalities than the same problems in a car. "
Pretty sure I'd take your money on that; aircraft are *very* complex and most every flight has a "malfunction" or two; window screen doesn't work, etc.
So now you'll have to redefine what you mean until it becomes a circular argument.
You may be uncomfortable flying, but facts are, well, facts.

Michael C.

Mark, you make an important point in your last sentence. I feel my fear of flying is rational, because all deaths are not equal. I would much, much rather die in a car crash than in a plane crash.

Per-mile fatality statistics are *not* the only factor at play!

Visitor

As Robert discovered, mechanical failures are not necessarily injurious, especially with redundant systems. When we harness fantastic power, like a jetliner's turbine engine, they (very rarely) fail dramatically. Transportation accidents are usually the drivers' fault. This week especially, trusting some highly trained guys to fly you six miles above the snowy midwest was a much better bet than driving. We arise each morning and take our chances afresh, and in the end we are all dead... I suspect Robert likes to feel he has control of his destiny, and the airlines do not offer so much of that. But he can build his own airplane and fly it on his own terms - thousands of people do!

Mark Uzick

"The questions I would pose to determine the safety of flying vs. driving would be: What percentage of mechanical malfunctions in cars result in fatalities?"

That question only helps to determine the relative danger of death from a flying mishap vs. a driving mishap.

The comparison of the safety of trips of equal distance, by different forms of transportation, must also consider the frequency of mishaps: where the danger of death during a trip is equal to the mortal danger of an average mishap times the expected frequency of mishaps.

Compared with driving, you should be more afraid of a flying mishap, but less afraid of a flying trip of equal distance.

What I think you might be getting at is that death from a flying mishap is likely to be so much more of a drawn out ordeal than from a driving mishap that we are more afraid of the far safer option.

I think it makes sense that people would prefer to take a chance of a sudden death in preference to a much lower chance of a horrifying death.

© Copyright 2013 Liberty Foundation. All rights reserved.



Opinions expressed in Liberty are those of the authors and not necessarily those of the Liberty Foundation.

All letters to the editor are assumed to be for publication unless otherwise indicated.