"Ten safest airlines"

[Beanbag]

The Economist reports on a ranking of 'safest' airlines, based on more factors than just historic accident record: Airline safety: Safe havens | The Economist (http://www.economist.com/blogs/gulliver/2011/08/airline-safety?fsrc=nlw|gul|09-06-11|gulliver)

What do PPRuners think of the list? There's one regular poster I can think of who would disagree with the inclusion of AMR, and some of the posters on the AF447 thread might have something to say about AF being on there.

[Wannabe Flyer]

I am assuming air accidents are not part of the criteria?

[pitotheat]

Given the number of accidents and incidents that have occurred with AF in the last 10-15 years how can this outfit be at the top of the list? This report deserves to be consigned to the bin along with reports on European Bank Stress Testing results, surveys that UK education is improving the life chances of our young because of rising exam results and the various hogwash out of Brussels.

[mogas-82]

These dimwits, probaly a news agency read a press release and said - yeah - now we can fill some voids in our online papers. The website (atra.aero) looks cheap and was registred less than one year ago and so does not look very credibility assuring (no info about the people). Seriously - AA is among the top ten???:ugh::ugh::ugh:It appears to me that they just wrote down the names of the airlines that came first to mind. I think it might actually be someone who just made something up to show the press how dumb their journalists are. Just read the disclaimer on atra.aero

EDIT:apparently they do have an address

[u674087]

Based on this criteria, which I don't believe has much a bearing on safety, where is Emirates and Qatar Airways?

[WHBM]

I can only conclude that this nonsense was written by the advertising department at The Economst, with a list of the ten airlines they will be pursuing soon for putting ad space in the Economist.

Incidentally, The Economist has produced a lot of comparable articles in recent years. Which is why I gave up reading it a good while ago.

[decurion]

The rating is done by a company called ATRA. They claim that that can rate the safety of an airline by looking at 15 factors (see ATRA HOLISTIC SAFETY RATING - Official website of the Air Transport Rating Agency - Geneva (http://www.atra.aero/ATRA_HOLISTIC_SAFETY_RATING-cms4.htm))

A large number of their factors (e.g. financial result, fleet age) have been analysed in many others studies, showing that there is no correlation with flight safety at all. The level of flight safety is determined by many factors which often can interact. I have strong doubts about the quality of the rating done by the company.

On the ATRA website you will find the term “aircrafts”. Makes you wonder how good this company is if they can’t even spell the plural of aircraft correctly.

[SLF3]

Have you read the article? It strikes a sceptical note (don't understand the statistics) and ends by saying passengers will remember the accidents...

[Pax Vobiscum]

Producing a useful comparison of airline safety is very difficult for a number of reasons:

1) Air travel is inherently very, very safe (for which the airline industry and their regulators deserve a lot of credit). So disasters are very rare events, particularly for an individual airline. If you want to produce statistically significant figures for such events, you have to go back a very long way into history to gather enough data - and I personally don't think 15 years is long enough. But even over 15 years, airlines change significantly. The people and machines are different, the routes may be different. So you may, in effect, be saying we think airline A was less safe than airline B 15 years ago, but you don't know whether that remains true today or tomorrow.

2) Sadly, there will be a next accident. When that happens the airline affected is suddenly going to appear much less safe. Air France have experienced more than their fair share recently, but whether that's bad luck or an indicator of something 'less safe' about their operations is a very difficult call to make. If things had turned out slightly differently or if their pilots had been slightly less skillful, either BA38 or US1549 might have involved heavy losses of life - and they might disappear from ATRA's list.

3) But neither of these two events could reasonably be considered to be related to the airlines involved. In BA's case, a previously unknown engine fault brought the plane down; for US Airways it was a bird strike, which could happen to anyone - so should these count in our statistics?

4) A further difficulty in comparing airlines is their route structure. If an airline flies to many third world destinations where airports and their facilities are less sophisticated, it's probably at greater risk of an accident than an equivalent operator in the continental USA - but that doesn't mean it's less safe. If you want to fly from NYC to LA, you probably have a choice of half a dozen airlines, but I'm willing to bet they're all very safe and your choice is going to be dictated by factors such as comfort, convenience and price. If you need to fly to Port Moresby or Timbuktu, you're probably going to end up with a final leg on a local operator with older, smaller aircraft. It's probably less safe than an equivalent journey across Europe, but if you don't want to run that (very small) risk, the answer is not to go there!

[alf5071h]

Pax Vobiscum, you gloss over a significant issue with BA38 or US1549.
Whilst the accidents were not of the operator’s making, the outcome may have been significantly worse with an operator who did not select, train, and support their crews in achievement of professional excellence.
Neither of the accident reports covered these aspects in detail, thus the premise cannot be proven, but a brief exploration of the facts behind the media ‘hero’ labels might reveal important aspects.

These accident scenarios and crew performance have similarities with modern safety concepts such as ‘resilience’ and how to avoid ‘the big one’, thus these accidents should be considered in any analysis.
Yes accidents will very rarely continue to occur, but skilful safety management (safety culture) might ensure an appropriate level of resilience and mitigation to lessen the outcome – as demonstrated by BA38 or US1549.
These aspects could be better indicators of relative ‘safety’ in an already safe transport system.

Resilience Engineering (http://www.resilience-engineering.org/)
resilienceengineering - erikhollnagel2 (http://sites.google.com/site/erikhollnagel2/resilienceengineering)
The paradoxes of almost totally safe transportation systems. (www.ida.liu.se/~eriho/SSCR/images/Amalberti%20_(2001).pdf)

[Pax Vobiscum]

I entirely agree (FWIW from my perspective back in the cabin :O), alf5071h, and I'm sure the flightcrews involved would be the first to acknowledge that they were at the pointy end of a large team and had been the recipients of excellent training and procedures. Despite this, it is not too difficult to envisage very similar circumstances in which no amount of skill and training could have avoided substantial loss of life and yet the airline might easily have played the role only of a helpless bystander.

Thanks for the links, which make fascinating reading. The last one has a superfluous ')' on the end, so if you get a 404 error, just delete the last character and it should work.