Aviation Accident Survival Rates — Lies, Damn Lies and Statistics

USA Today recently published (9th August 2013) an interesting story by Mr Gary Stroller concerning commercial aviation accident survival statistics.   You can find the story on their web site at:


I think this is an interesting line of inquiry, but I don’t believe that the analysis presented addresses the whole picture and so can be misleading.  The main point of the story is made in the opening sentences of the article, namely:

 “More passengers are surviving fatal airline accidents during landing, but survival rates during stages of flight when most accidents happen have not improved from decades ago, a new USA TODAY study shows.”

 I am quoted in the article as follows,

“Edwin Galea, a professor at the University of Greenwich in England who is an aircraft evacuation expert, said the USA TODAY study is “interesting” but omits some serious accidents without fatalities.”

However, this is only part of the concerns I conveyed to Mr Stroller.  I have undertaken quite a bit of research on evacuation from aircraft including analysis of past accidents and appreciate how difficult it is to come to any meaningful conclusions when analysing aviation survival statistics (see publication list).  While I do not doubt the accuracy of the data presented in the article, I believe that the analysis is potentially misleading.  You have to look at the questions being asked, not just the answers being presented.  I do not think the questions that are being posed in the article are quite right.  My quotation in the article covers just one of my two main concerns.  These are:

1)      The USA Today analysis covers aircraft in which at least one person has died.  So the analysis definitely includes what could be considered to be serious accidents.  However, this is just a sub-set of serious accidents as it does it does not include serious accidents in which no one died, take for example just three recent accidents:

  1. US Flight 1549 on 15 Jan 2009, an A320 out of LaGuardia heading to Seattle, carrying 150 pax and 5 crew.  Aircraft was written-off after a forced landing on the Hudson River, there were five injuries, two of which were serious, yet everyone survived.
  2. BA Flight 38 on 17 Jan 2008, a B777 coming in to land at Heathrow, carrying 136 paxs and 16 crew.  Aircraft was written-off, there were 47 injuries, one of which was serious, yet everyone survived.
  3. Air France Flight 358 on 2 Aug 2005, an A340 coming into land at Toronto, carrying 309 paxs and 12 crew.  There was a serious fire and the aircraft was written-off, there were 12 injuries, yet everyone survived.

By anyone’s reckoning, these were three very serious accidents — so much so that in each case the aircraft was written-off – yet no one died.   The USA Today analysis simply ignores these type of incidents and so incorporates a bias in the data.

2)      The USA Today analysis includes accidents in which everyone on board died.  Thus the analysis includes technically non-survivable accidents.  These are accidents in which, given that the incident occurs, no feasible technological development is likely to improve the chances of survival as the trauma associated with the accident is simply too severe, for example:

  1. Mid-air explosion — Pan Am Flight 103 (Lockerbie), 21 Dec 1988, B747, all 243 paxs and 16 crew killed.
  2. Uncontrolled Flight into Terrain — Swiss Air Flight 111, 2 Sept 1998, MD11, all 215 pax and 14 crew killed.

The industry addresses these type of problems by introducing technological or procedural improvements to prevent them from occurring in the first place — prevention rather than mitigation.  The issue here is reducing the frequency of these events from happening in the first place.

Mixing technically survivable and non-survivable accidents into the discussion on survivability in aviation accidents confuses the issue as advances in technology are not likely to impact survivability in non-survivable accidents.  A clearer understanding of whether or not we are improving survivability in aviation incidents would be derived by splitting the discussion into two parts:

i)        First, consider survivability in technically survivable accidents.  The purpose of this investigation is to establish whether or not your chances of surviving a survivable crash is getting better or worse as a result of the advances in technology and regulation.  Related to this would be determining how likely you are to be involved in a survivable crash and whether or not this is improving.

ii)      Secondly, consider the frequency of technically non-survivable accidents.  The purpose of this investigation is to establish how likely you are to be involved in a non-survivable crash and whether or not advances in technology and regulation are improving these chances.

It is essential to filter the raw aviation accident data in an appropriate and meaningful way if truly meaningful conclusions are to be drawn as to the current state of aviation safety and whether or not it is improving.


1)       “Aircraft Accident Statistics and Knowledge Database: Analyzing Passenger Behaviour in Aviation Accidents.”, E.R.Galea, K.M.Finney, A.J.P.Dixon, A.Siddiqui and D.P.Cooney. AIAA Journal of Aircraft, Vol 43, Number 5, pp 1272-1281, 2006.

2)      “An analysis of exit availability, exit usage and passenger exit selection behaviour exhibited during actual aviation accidents.”, E.R.Galea, K.M.Finney, A.J.P.Dixon, A.Siddiqui and D.P.Cooney. The Aeronautical Journal of the Royal Aeronautical Society, Vol 110, Number 1106, pp 239-248, 2006

3)       “An analysis of the passenger to cabin crew ratio and exit reliability based on past survivable aviation accidents.”, E.R.Galea, K.M.Finney, A.J.P.Dixon, A.Siddiqui and D.P.Cooney. Human Factors and Aerospace Safety, Ashgate publishing, 5(3), pp 239-256, 2005.

4)      “An Analysis of human behaviour during aircraft evacuation situations using the AASK V3.0 database.”. Galea E.R., Finney, K.M., Dixon, A.J.P., Siddiqui A., and Cooney D.P. The Aeronautical Journal, Vol 107, Number 1070, pp 219-231, 2003.

5)       A Database to Record Human Experience of Evacuation in Aviation Accidents.  The Aircraft Statistics and Knowledge Database (AASK). Authors: E R Galea, K M Finney, A J P Dixon, A Siddiqui and D P Cooney.ISBN 9780117908390, CAA PAPER 2006/01. June 2008.


6)       The AASK Database: Aircraft Accident Statistics and Knowledge. Authors: E R Galea, D Cooney, A Dixon, K Finney and A Siddiqui. UK CAA Paper 2002/3 ISBN 0 86039 803 X, 2002.


If you enjoyed this post, make sure you subscribe to my RSS feed!
This entry was posted in aircraft crash, Aircraft Evacuation, Evacaution, Fire, survivability and tagged , , , , . Bookmark the permalink.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.