The AirSafe Journal - Issue 11

Airline Rankings and Risk Decision Making
27 September 1997


This issue of the AirSafe Journal introduces an improved airline fatal event listing that includes regional rankings by fatal event rate per million flights. Also, the author provides an example of a method to determine the safety of an airline or an airline flight.


Airline Rankings by Fatal Event Rate

The previous issue of the AirSafe Journal was devoted to a critique of the Air Travelers Association "Airline Report Card" which included a safety rating of 260 airlines around the world. I criticized that publication in part because the formula used to rank airlines contained a number of biases. My most significant disagreement was the implication that a single numerical measure of risk, namely the rate of fatal accidents implies that an airline either passes or fails with respect to being sufficiently safe. In my experience, there is no one measure or even a combination of numerical measurements that can determine whether an airline is safe. As I have stated in the first issue of the Journal, what a person considers to be safe depends on both the level of risk and the context in which that risk occurs.

In the area of rating airline safety based on a risk measure such as fatal events, the airline industry and its government regulators have been generally against the idea. For example, in reference to the Air Travelers Association ratings Cable News Network quoted FAA spokesperson Les Dorr as saying that "Fatal accidents are such infrequent events, that for statistical purposes of rating airline safety, it just doesn't work." I completely agree with this view. I also agree with former FAA associate administrator Anthony Broderick who in the publication Aviation Safety & Maintenance disagreed with the concept of the Air Travelers Association rankings and stated that any such ranking system must "generate results which truly distinguish more risky choices from less risky choices."

Currently, airline risk rankings are not routinely published by the FAA or any other government agency in the U.S. or any other country. I do not know the reasons why such rankings are not developed by these agencies, but I do know that the general public is interested in this information. Based in part on the many questions that my site's visitors have sent me, there is a strong desire by the general public for such information. I also believe that publication of such information will not lead to panic on the part of the traveling public. I also believe that an airline's historical rate of safety related events is one of the measures that a person should consider when making a decision to fly. I saw no reason to keep such information from the public, so I decided to add information on fatal event rates by airline to my site for many of the major airlines of the world.

The site contains the fatal event rate since 1970 for those listed airlines as well as the number of fatal events, the year of the most recent fatal event, the estimated number of flights during the time period, and links to lists of those airlines that had at least one fatal event. The following section takes this information a step further by describing my own rating scale that I apply to airlines that I fly.

Back to Contents for this Issue

A Personal Airline Safety Decision Method

In the last Journal issue, I criticized the rating system of the Air Travelers Association in part because the method used did not allow for the use of someone's judgment. Criticizing that method or any other safety rating method is easy. It is much more difficult to create an alternative rating system that would be superior. I have struggled with this issue for quite some time and eventually came to the conclusion that I am unable to create a superior system. However, I do believe that a superior method is one that is base on an individual's values with respect to safety with respect to flying on a particular airline flight. I will not even pretend that I can determine another person's values in this matter. However, I can speak with some authority on what my values are in this situation. In my case, safety is determined by a combination of objective and subjective information and that any judgment that I make is subject to change by any number of factors. Some of those factors and conditions follow:

  • Airline size
  • Kind of flight (scheduled or unscheduled
  • Airline history
  • Airline management
  • Airline reputation
  • Airline home country
  • Airline safety record
  • Aircraft safety record
  • Aircraft capability
  • Regulatory environment
  • Airline cockpit, cabin, and maintenance selection and training
  • Route of flight
  • Weather
  • Airports used
  • Size of aircraft
  • Aircraft history

This is not an exhaustive list. I would not consider all the factors all the time, in part because the information may not be available or if available it may not be very detailed or very reliable. Also the importance of an individual factor may vary greatly depending on the context of my decision. For example, if I want to travel by air within a particular country, I will only be able to choose between those airlines allowed to fly domestically within that country. The airline I choose may have the best safety record of any airline in that country, but a much worse record than any airline in my home country. My concern in this situation is to choose a flight on an airline that I believe is is less risky than the available alternatives.

The airline safety record is only on factor, and even that factor has many components. I frequently refer to the passenger fatal event rate in this site because the risk of death is the risk that is of greatest concern of most passengers. Fatal events are also the kind of safety related event that is most likely to be reliably reported by either the media or by civil aviation authorities. However, when I make a decision on whether to take a particular flight or use a particular airline, I don't make my decision based solely on the fatal event rate or any other single measure. I do so by first taking into account a wide range of information and then by considering the following questions:

  1. Would I allow my minor child to fly unaccompanied?
  2. Would I allow my minor child to fly with an adult?
  3. Would I allow an adult family member to fly?
  4. Would I allow myself to fly?
  5. Would I allow myself to fly only under special circumstances?
  6. Would I allow someone I know on the flight?
  7. Would I allow someone I did not know on the flight?

My personal interpretation of the level of safety of a particular flight or airline would depend on how I answered the question. For example, if I answered the first question yes, then all the other questions below it would also be yes. This is neither a comprehensive set of questions or a set of questions which can deal with everyone's safety issues. This is merely my personal scale for rating the safety of an airline or an airline flight (I'll call it the Curtis Criteria). This method may not work for many people, but it works well for me.

Back to Contents for this Issue
The AirSafe Journal - Issue 11 -- Revised: 24 May 2015