American Scientist

To the Editors:

The article "Flying and Driving After the September 11 Attacks" (Macroscope, January–February), by Michael Sivak and Michael J. Flannagan, is a more effective promotion of airline safety (both before and after 9/11) than anything that could have been written by a paid airline lobbyist. It repeats the old saw that flying is many times safer than driving. However, while the article's conclusions are acceptable in the broad sense, they are highly skewed by the common definition of safety in terms of passenger-miles per fatality. This is reasonable from the point of view of those involved in the economics and operations of flying machines. However, as a human (and pilot) with a finite life span, I should point out that the numbers on gravestones represent time, not miles traveled. On an odometer basis, my perambulations around the hearthrug by rocking chair are infinitely more dangerous than an astronaut's wildest rides through space.

The pertinent question for a human should be: "Is it safer to spend an hour in a car or in a commercial aircraft?" Since the distance traveled per hour in a commercial jet is approximately 10 times that in a car, the relative merits of air travel compared to car travel diminish dramatically on a time basis, to the point where the safety per hour for trips of around 300 kilometers is probably equal for both forms of transport. When the additional airport delays involving several hours spent going nowhere are included (equivalent to at least 100 kilometers of car travel), that distance could be increased considerably. Obviously, it would not significantly affect longer nonstop and transoceanic flights.

It is unfortunate that comparisons of travel safety by ship/train/car/aircraft are normally made on a mileage basis. It would be very revealing to see such comparisons based on time. Insurance premiums provide a reasonably objective assessment of relative risk. For a private pilot, the annual liability insurance for a single-engine four-place aircraft is similar to that for a mid-price SUV. On a mileage basis, the safety of the aircraft may be greater than that of the car, but in terms of elapsed time the insurance is many times more costly.

In the end, perhaps the most important question is the one that was plastered on railway stations in London during the dark days of the WWII blitz: "Is your journey really necessary?"

Peter Jones
Calgary, Alberta, Canada

To the Editors:

While reading the article about the relative safety of air and auto travel, several questions came to mind. I wonder how much of the reduced risk associated with commercial air transportation results from professional transportation providers rather than something inherent in flying.

Specifically, I wonder if the risk associated with traveling by commercial bus or commercial train is similar to that of commercial airlines. Considering general aviation might shed some light on the link between professionalism and risk in transportation. Unlike most motor vehicle operators, who get their driver's licenses when they are young and renew them without question for the rest of their lives, general aviation pilots must be medically certified and are required routinely to demonstrate competency. I suspect that the risk associated with general aviation lies between those of commercial aviation and driving (and probably closer to that of commercial aviation).

I also wonder how much of the risk associated with driving is attributable to other vehicles on the road. In this regard, flying would seem to have a definite advantage over driving.

J. P. Mellor
Rose-Hulman Institute of Technology
Terre Haute, Indiana

To the Editors:

The central point of Sivak and Flannagan's article is on target—flying is generally much safer than driving. An important contributor to the higher risk in driving is that U.S. road-safety policy has inappropriately devoted so much focus on increasing survivability when crashes occur, as described in my article "Traffic Crashes" in the May–June 2002 issue. Airline safety has improved so dramatically by correctly focusing on preventing, not surviving, crashes.

I do think the article overstates the admittedly large differences in risk between flying and driving. For trips of distances for which there is a reasonable choice in mode of transport, the flying is likely to be on small aircraft operated by commuter airlines. These aircraft, and airlines, have higher fatality risks than large jets flown by major airlines. Commuter airline casualties were excluded from the fly-versus-drive comparison.

It is also worthwhile to consider that all passengers on an airline flight have near identical risks, whereas driving risks vary enormously between drivers. A typical driver killed is a drunk unbelted 19-year-old male driving at illegal speeds an hour or so after midnight. Typical airline passengers (and typical American Scientist readers) have personal profiles markedly different from those of drivers killed in traffic, and accordingly have far lower than average driving risk. Risks in driving, and also in flying, are however substantially higher than many other risks (tornados, chemicals, nuclear power) that attract much attention and resources.

Leonard Evans
Science Serving Society
Bloomfield Hills, Michigan

To the Editors:

Studies show that risky or inattentive driving behavior is a factor in most fatal motor vehicle crashes. Examples include driving too fast for conditions, driving while impaired (alcohol, drugs or fatigue) and failure to buckle the seat belt.

Travelers want to know how safe airline travel is relative to driving responsibly. To answer this question, Sivak and Flannagan would have to purge their input data of accidents caused by well-known and easily avoidable risky behavior.

Robert R. Piper
Berkeley, California

To the Editors:

The "Macroscope" article carefully analyzes the risks of flying versus driving, showing that the personal risk difference does not explain or justify the recent 18 percent drop in public flying.

The explanation is plainly, although unintentionally, given in the subsequent article (Marginalia) by Roald Hoffmann: "? [W]hat matters is a heady mix of factors in which psychological attitudes figure prominently." Such as my attitude toward being searched and harassed by three idiotic airport inspectors in a row, who, in their frustration at not finding anything to steal, requested my passport to run through the x-ray machine!

Werner Barasch
Los Gatos, California

To the Editors:

Sivak and Flannagan consider the relative risks of flying versus driving, coming to the oft-quoted conclusion that flying is safer. I have two questions regarding their analysis.

It is often stated that the majority of automobile fatalities occur within 25 miles of home (a distance that is normally not flown by major airlines). Drivers who live in rural areas frequently use the interstates to make trips within this range. Was any attempt made to exclude these "local" fatalities?

Secondly, it seems to me that it is more relevant to compare fatalities per trip, rather than fatalities per kilometer. The authors make such a comparison based on the conventional assumption of a linear dependence of driving-risk on distance. What are the data supporting this assumption, and how were they analyzed?

James H. Taylor Central Missouri State University
Warrensburg