An interesting old thread on the sr111 board. There are 4 participants -Mark, myself, an unnamed poster, and Paxsafetyguy who is a highly respected Canadian safety expert:

It began with the following article:

NTSB RELEASES AVIATION ACCIDENT STATISTICS FOR 2001

Washington D.C. - Aviation accident statistics released today by The National Transportation Safety Board show a decline in the scheduled
U.S. airline accident rate in 2001.

In 2001 there were 36 accidents on U.S.scheduled airlines, including the 4 crashes of September 11. Because the crashes of September 11, 2001 were the results of criminal activity, those crashes are included in the totals for scheduled U.S. airline accidents and fatalities, but
are not used for the purpose of accident rate computation. The remaining 32 accidents in 2001 result in an accident rate of .317 per 100,000 departures. These numbers represent a decrease from 2000, when 51 accidents were reported for a rate of .463 accidents per 100,000 departures.

The 531 fatalities associated with crashes involving U.S. scheduled airliners last year is the highest total since 1977, when two jumbo
jets collided in the Canary Islands. Half of last year's fatalities - 265 - occurred aboard the four hijacked airliners on September 11.
Other than a ground worker who was struck by a propeller at an airport in August, the remaining fatalities (265) occurred when American
Airlines flight 587 crashed in New York on November 12.

Accident rates for both scheduled and non-scheduled 14 CFR part 135 service decreased in 2001. The scheduled service rate decreased from
1.965 accidents per 100,000 departures in 2000 to 1.407 in 2001. For unscheduled, on-demand air taxis, the rate decreased from 2.28 to 2.12
per 100,000 flight hours. Table 9a reflects the effect of the FAA revision to activity estimates on accident rates for on-demand air taxis from 1992-2000.

Despite reporting fewer accidents in 2001, the accident rate for general aviation aircraft increased slightly from 6.33 accidents per
100,000 flight hours in 2000 to 6.56 accidents in 2001. General aviation was the only category if air transportation to report an increase in its accident rate, which is attributable to the fact that less hours were flown by general aviation aircraft in 2001 than in 2000.

Table 12 showing Air Carrier Occurrences Involving Illegal Acts has been added to the 2001 aviation statistics. The table outlines
accidents caused by sabotage, suicide or terrorism from 1982-2001. As with 2001 statistics, accidents caused by illegal acts in previous years were not used when calculating the accident rate for that year.

Tables 1-12 providing additional statistics are available at
http://www.ntsb.gov/aviation/Stats.htm.

Paxsafetyguy:

I find it hard to understand the logic being used by the NTSB in excluding the four 9-11 crashes from the 'accident' rate computation. All four events represented one of more failures in the aviation safety net. By excluding this data, the NTSB is deceiving the travelling public about the true safety of aviation. The crew and passengers in those aircraft are just as dead as any other victims. How can passengers, who may have a choice between car, bus, train or even ships, make an informed decision if we are, as it were, dealing
some of the cards out of the deck. Can anyone explain this to me?

My reply:

Paxsafetyguy, I absolutely agree with you 100%. That is outrageous. Obviously the lax security system that the airlines were using shares
a large part of the responsibility for those 4 crashes. It's time they were more honest with the statistics they present to the flying
public. Many family members were deeply offended in 1998 when sr111 was excluded from the accident statistics considering that most of
the passengers were American, the plane was manufactured in this country, etc. In my mind these figures are deceptive to say the least.

I posted the following article:

Helicopter engineer cleared of manslaughter after crash

An aircraft engineer has been cleared of three charges of manslaughter after a helicopter crashed to the ground killing everyone on board.

Paul Kenward, 47, from Biggin Hill, Kent, was cleared at Oxford Crown Court after the accident involving a Hughes 269C helicopter.

The aircraft plummeted to the ground on March 8, 2000, at Twyford, Berkshire, when the tail boom broke away and became entangled in the
rotor blades.

Kenward had ignored manufacturers' instructions to replace a cracked part which supported the tail boom and had it welded instead.

Brendan Loft, 38, Dennis Kenyon, 18, and Jane Biddolph, 23, all died instantly when the helicopter smashed into the ground and burst into
flames minutes after taking off on a pleasure trip from the Booker airfield in High Wycombe, Bucks.

Pilot Mr Loft of Reigate, Surrey, and his companions, from Shoreham in West Sussex, were pronounced dead at the scene of the crash.

Kenward, a trained aircraft engineer, licensed by the Civil Aviation Authority and with years of experience, pleaded guilty before the
trial to three separate charges in connection with the fatal accident.

He admitted allowing an aircraft to fly without a valid certificate of authority, a further charge of endangering an aircraft and failing to
make an entry into an aircraft log book. Sentencing has been adjourned for reports and will be held at a date to be fixed.

The judge Mr Justice Aitkens told Kenward that he was considering all possibilities including a term of imprisonment. The maximum sentence
Kenward can face is an unlimited fine, two years in prison or both.

Outside the court, Brendan Loft's father, Walter, 74, said: "We are very disappointed with the result. I feel terribly frustrated with the
court system. This verdict sends out the wrong message to those involved in the maintenance of aircraft. We have lost a wonderful, wonderful son."

Story filed: 16:39 Tuesday 26th March 2002

A poster says (regarding statistics article)

They are apparently differentiating between an accident and a deliberate act.

Paxsafetyguy responds to helicopter tragedy article:

This wretched helicopter 'accident' - really no accident at all - makes my point about the NTSB statistics. When an engineer/mechanic
knowingly installs bogus parts, fails to do a required repair, installs a defective IFES, busts weather minima, overloads an aircraft, etc, etc, etc, how is that materially different from a terrorist act? Again, the poor victims had no idea this flight would be there last, they had no control over it and they ended up just as dead as those in the -- and I gag when I say this - "normal accident."

So now we have two sets of books, the 'normal' and the 'non-normal' which must be added together to get an accurate picture of the actual state of safety. This is the sort of duplicitous bookkeeping that gave us Enron and other fiascos.

Aviation safety should mean exactly that - people get on an airplane alive and in one piece - and they get off the airplane the same way.

For shame, for shame.

Poster responds to Paxsafetyguy�s statistics comments:

I must disagree.

It is easy to blame "the system" for the events of 9/11. However, the fact is that it was not so much a failure of "the system", but rather people who knew how to exploit it.

First of all we have the box cutters, knives, or whatever it was that was used by the hijackers. The truth is that those items were perfectly legal to have on board. It seems like I remember reading somewhere that you could have a knife with up to a 4 inch blade. Now, whether or not that should have been legal could be debated, but it was legal.

Next is the issue of the flight training the hijackers took. It is certainly not unusual for a foreign national to study in the US,
flight training or otherwise. It happens all the time.

If there was a failure, it was a failure in intelligence to identify the threat and identify poeple with obvious terorist connections. It
was also a failure in that some of the hijackers were in the country on expired visas. However, it is certainly not up to the FAA to
enforce immigration law.

I don't believe the NTSB was trying to deceive anybody. I believe they were presenting the facts as they had them, and I agree with
their decision about not classifying the events of 9/11 as an accident.

Hindsight is always 20/20. Looking back, we can see where some red flags maybe should have been raised, but this attack was without precedent. Be honest, before 9/11, would you have ever considered that a foreign national taking flight training in the US was doing so for the purpose of crashing an airplane into a building? I wouldn't have.

Paxsafetyguy:

It's your right to disagree, and I am sure others would too - the FAA, the airlines, the airports, etc. They have powerful interests
to protect. But your argument fails to convince me. I don't find it logical. For many years, aviation professionals have winced at the
word 'accident' just as they have at 'pilot error.' (For years, the USN has used 'mishap' for this very reason) Accident and pilot error
cover a multitude of sins, and were convenient cop-outs. But hiding behind facile rationalizations only masks the true state of the situation and dresses a six up as a nine. A failure in any part of the system is a failure of the system. Where do we stop engineering
these statistical "cut-outs" and exceptions. We are only kidding ourselves - and who for? - the living who should be held accountable,
or the dead, who have every right to hold others accountable, but who rely on us to do it for them. I will not have full faith in aviation
safety until those in authority look at the big picture and uses precise nomenclature.

Final point - not to put El Al on a pedestal - they lost a 747 over Amsterdam - but why should their superb security go statistically unrecognized - which is the upshot of what you are saying - when those with lax security standards can slip by unnoticed, as it were,
and don't have to be held to statistical account?

Poster answers:

I don't understand. What are you talking about with El Al and statistics?

Paxsafetyguy responds:

Sorry if my post was a bit cryptic and the reply delayed - I have to work!

My point was that some companies/countries go to great length to deal with security - others have not. Why should those that do not go the
extra mile get a free ride (by having their losses not included, i.e. not counted against them) in the national accident rate? Aviation
safety should be end-to-end and objective. If we start making exceptions on the basis of mens rea - the guilty mind - then should
be start excluding accidents caused by deliberately doing other criminal or illegal acts - like busting weather minima, installing
bogus parts, etc. Why do we have security in the first place if it is not to prevent this sort of thing? And even if the standards were inadequate in one country (or many) is that also an excuse to exclude a loss? I don't think so. Does DoT discount driver deaths that were suicide? I don't think so. Bus accidents where there was illegal interference? I don't know, but I don't think they should. Where else are we going to capture not only the losses, but the trends?

Hope this offers some clarity.

Poster response to Paxsafetyguy:

I believe our differences lie in our definitions of what is an accident and what is not. I'm going by the "dictionary" definition,
meaning that I believe an accident is something that was unintended. Despite all of the mistakes that may have contributed to SR-111, I
don't think anybody was intentionally trying to crash the airplane. That crash was definitely avoidable, but it was not intentional.

The attacks of 9/11 were intentional, and that is where I, and apparently also the NTSB, draw the distinction.

Paxsafetyguy:

I appreciate the engagement. And I agree, the issue seems to hinge on the question of whether there was a deliberate intention to do harm
to others. I believe the SR111 type of event (and many others were there where the safety of people were deliberately compromised, usually for short term financial goals and the terrorist events is merely a question of degree, my main point is, from the point of aviation safety, the integrity of the system and the perspective of
the victims (which is what the safety system has been set up to avoid) the difference is moot. Moreover, there seems to be altogether
too much of this stuff - KAL 007, Air India 181/182, UTA's DC10 that was blown over Africa; Iran Air Airbus that the USS Vincennes
dispatched, EgyptAir 990 and the Silkair 737 - and those are only some of the recent examples. Why not include them in the rate - spurring greater efforts.

Not doing so suggests that aviation safety is a matter of mens rea. I don't think this holds up to logical analysis:

If you accept the notion that intentions are followed by acts are followed by results. You can connect the following sequential pairs
with arrows that flow straight from left to right, or diagonally from left to right:

Good intent(GI) Good Act(GA) Good Results(GR)
Bad intent(BI) Bad Act(BA) Bad results(BR)

Then GI>GA and BI>BA are deliberate acts, the first professional, the second criminal;
BI>GA is a thwarted crime; GI>BA is to yield to temptation; Both GA>GR and BA>BR are desired events; and GA>BR is your "classical accident"; BA>GR is serendipity (of which there are many.)

The whole safety and regulatory structure has been put in place to thwart BA, whether they are the result of GI or BI. Success is
measured (my thesis, anyway) by GR or BR, not by GI or BI.

Why should passengers be have to worry about "two sets of statistical books" - the second of which they have to calculate themselves? I submit it would be virtually impossible for the average person to do
this. I keep hearing the endless mantra of "Data Driven" approaches -(which contains its own logical fallacy, but that is a horse of
another garage), but that but when the data is not as flattering as it could be, it is conveniently swept under the rug. I don't like it.

Pax

Mark responds to Paxsafetyguy:

Pax,

I used to be a "statistician" for a company that did executive compensation studies. Typically our clients were interested in justifying very lucrative compensation plans for top executives. We collected data from other comparable companies. We then correlated various elements of company performance with the compensation. It wasn't very hard to find a set of independent variables that would show that the comp plans were not only reasonable, but consistent with the practices of industry peers. Of course
any real statistician will tell you that this is an outrageous abuse of statistical technique: no experimental design, no controls, blatant
disregard for the principles of statistical inference. Fortunately, I am no longer in the business of statistical quackery, but I still know it when I see it.

My favorite aviation safety measurement is deaths per passenger mile. As a member of the flying public, what I really want to know is, what are my chances of getting there alive? I'm really not that interested in my chances
of getting killed in a given mile. DPM might make sense if I was trying to compare the relative safety of driving versus flying, but if I had to drive 3000 miles to Seattle, I wouldn't go. And you certainly can't drive from New
York to Geneva. For my purpose, deaths per flight or flight segment would be a much better measure of safety. Even deaths per flight hour would be better than DPM. I think it's clear why DPM is preferred: because it's the most
favorable (safe) measure.

From the same perspective, what sense does it make to tell a passenger that the likelihood of getting killed _unintentionally_ is x%? The intentionally killed are just as dead. Why bury the truth in the fine print? Once again,
one can't help but wonder whether the purveyors of the "statistics" aren't using the desired outcome to choose the measurement.

One would think that safety advocates would favor the most conservative measure of risk, whereas air transportation businesses might argue that terrorists and suicidal pilots shouldn't be counted against them (not
convincing, but understandable). This sort of thing leads me to believe that the influence of the industry is far greater than that of safety advocates. Is there really any doubt?

Interesting too that, in the U.S., if you drive drunk and kill someone, you can be charged with murder, but if you deliberately side-step certification, install shoddy equipment, neglect maintenance and otherwise knowingly
disregard aircraft passenger safety, and the result is an air crash, you are essentially protected from both criminal and civil liability. I guess drunk drivers need an industry association. Perhaps the ATA charter would provide a few ideas. Maybe they could convince the deportment of transportation to
alter the drunk driving death statistics to exclude those involving deliberate intoxication.

- Mark

Paxsafetyguy:

Mark:

Your post is both fascinating and chilling: you confirm what many of us laypeople have thought all along: that statistics offered as "proof" by an interested party rarely survive the test of
objective analysis. It is troubling enough that the general population seems indifferent or innumerate to such sleigh of hand. It is even more troubling when those who are responsible for our safety seem that way. Whenever I see a statistic, I start asking questions � exactly the types of questions that you cited: what
is the population size, how recent, what are the inclusions, what are the exclusions, the definitions, how was the data collected, who
collected it, etc, etc. It is funny how often a statistic, introduced with such authority, such certainty and such precision will crumble
or fall with even a mild degree of due diligence.

Of course, comparing the safety of various modes of travel is extremely problematic. The difference between jumping in the car to
drive to the local 7-11 to buy a loaf of bread, and taking a flight overseas is only the beginning. To provide a meaningful comparison,
we would have to take travelers going between identical city pairs, with a full choice of car, bus, train, (boat) and plane. Even if we
were able to do that, which would be impossible, we would run into the problem of which measure to use. Anything that measures time or
distance favors the plane, because it travels faster (i.e. covers a greater distance in the same time) than the other modes. There are
several interesting papers on the Internet to this effect. For this reason, I join you and many other aviation safety professionals in
prefering to measure safety by mishaps per flight (potential for death) and deaths per flight. Somehow we have to convince government
authorities to stop relying on deaths per hours � it is virtually meaningless.

In a related observation, commercial users of data are often a little more savvy with numbers than the average person. For this reason, I
found it particularly interesting that the insurance world pulled the pin on aircraft risks post 9-11, forcing the governments of the US,
Canada and Europe to pick up the slack. We were told the airlines were safe before 9-11 and again, after major initiatives, after 9-
11.

Of particular note in this country (Canada) the government yesterday introduced a $24 Cdn ($15 US) round trip security tax that is supposed to generate as much as $1B Cdn ($620M US) of profit over 5 years. I find this so incongruous with an earlier suggestion that they charge about $2 Cdn ($1.24 US) per flight to return the level of
safety at Canadian airports to what they were before. We might even get Rescue capability (no longer required); Response time of two
minutes, not to exceed three, to the end of the furthest runway (no longer required) and coverage of all turboprop and jet passenger
traffic (no longer required). A small portion of the security tax would fund everything. Nobody bothered to connect the dots and
incorporate safety with security. And so we await our next disaster.

Where does this leave us? With the onus on the authorities to be as open and honest and inclusive as possible with all statistics. I am
not knocking statistics � they can be a very powerful and useful tool, when used, as you point out, with a sense of professional
ethics. Even then, with all the imperfections in collecting and retrieving data, they are often flawed and are more probabilistic that factual. Unfortunately, this puts a large burden on the consumer to be a wise consumer of statistical data. Sadly, few people have the time or training to do anything other than accept what is given them.

Pax

Mark:

Here's an excerpt from an Article entitled Flight into Danger, 8/7/99,
From New Scientist magazine, vol 163 issue 2198

Andrew Weir's The Tombstone Imperative-The Truth About Air Safety is
available in Britain now (Simon & Schuster, �16.99) ... If you're interested
in the book, it can be found at the Amazon.co.uk:

http://www.amazon.co.uk/exec/obidos/ASIN/0743415833/qid=1017790823/sr=1-5/re
f=sr_1_3_5/026-2459069-7030807

Flight into danger 07 Aug 99

So you think a jet plane is the safest way to get where you want to go? It ain't necessarily so argues Andrew Weir

IT'S that time of year again-the season of mile-long check-in queues, mysteriously delayed takeoffs and wandering luggage. Every day, all over the world, tens of millions of us will be joining those queues.

So many passengers, so many planes. And with air travel growing by about 7 per cent every year, more are on the way. No wonder airports are barely able to keep up. But at least we can take some comfort from the airlines' assurances that the commercial jet is the safest form of transport ever invented, that flying is
as safe as technology can make it and getting ever safer. Can't we?

Unfortunately, few things in life are what they seem-and this is definitely the case with air travel. All right, so we may need some reassurance when we are stuck in a cramped aluminium tube surging at 900 kilometres per hour, 10 kilometres above Earth. But while working on a TV documentary series on air
crashes and researching my book, I discovered that the reassurances we're given are about as scientific as a belief in the curative powers of a rabbit's foot.Take the claims about flying being the safest form of transport. If you plot the number of fatal accidents against distance travelled, you end up with 0.03 deaths per 100 million kilometres for commercial aircraft versus 0.1 deaths per 100 million kilometres for rail travel.

What the airlines don't tell you is that this form of comparison effectively dilutes the accident rate for aircraft. Aircraft usually travel huge distances while cars and trains don't. And while the risk of having a fatal
accident in a car or train is spread more or less evenly across the journey time, the opposite is true for planes: 70 per cent of all aircraft accidents take place at takeoff and landing, which is only 4 per cent of journey time.

A better measure is to plot the number of deaths against the time travelled. This is fairer, since many car and train journeys last as long as plane journeys. But it still doesn't take into account the concentration of
accidents around takeoff and landing.

The most accurate method is to compare the number of deaths with the number of journeys made. So accurate, in fact, that this is the measure used by the industry and its insurers. This makes much more sense, because what matters
to the individual is the journey, not how long it took or how far it went. Also, it enables comparison of different types of jet, both long haul and short haul.

By this measure, air travel takes on a rather different complexion. Deaths per 100 million passenger journeys are, on average, 55 for airliners compared with 4.5 for cars, and 2.7 for trains. Only motorbikes, at 100
deaths per 100 million passenger journeys, are more risky than aircraft on this basis.

Mark:

Maybe statistics should have disclosure labelling requirements, like nutritional information on food or health warnings on tobacco and alcohol products. No, too complex.

But seriously, I'm not generally one to favor over-regulation, but I'll almost always support disclosure requirements. Barb had a good idea awhile ago. Airlines should be required to provide accident and incident statistics
when they issue tickets. And maybe a few other tidbits like aircraft age, mishaps for aircraft type etc.

FWIW,

- Mark

I was lead to believe after PanAm 103 the faa was going to make sure we were protected by enhancing airport security and going full steam ahead with checked baggage scanners? Considering the number of years between that crash and 911? I would have thought they (faa) would have had ample time.

Also, knowing jane garvey was a bigwig at Boston Logan Airport prior to her becoming the Administrator of the faa...I find it rather confusing and shocking to come to learn (the very hard way) that we were no better off all these years after PanAm 103 then we were before it.

I don't believe there was ever enough conversation about the lack of oversight in airport security post 911. The spin was mind boggling though. How can so many aircraft be hijacked in the same day <in the US>, IF the faa were on the job - as they said they were, post PanAm 103?

How is it that all those innocent human beings were dubbed heroes? I thought a hero was someone who died acting in a heroic way. On the battlefield (just one example).

I'm not sure any of those innocent human beings that went to work that day thought about giving their life for their country (as the spin went), but who knows.. I could be wrong. I'd just have to bet though, they all had plans that didn't include giving up their life that day.

Heroes.. beautiful sentiment after the fact, but IMHP? I'd bet anything they would have opted out of that honor.

IP whenever I think about 9/11 I always remember Victoria Cummock who lost her husband in the tragic Pan Am 103 crash. She worked very hard as a safety advocate to get airlines and the government to become aware of how urgent it was to improve security due to the obvious potential for terrorist acts. It seems to me that at one time Al Gore ( he headed an aviation committee as VP at that time) had promised to help her cause and then backed out at the last minute after the airlines lobbied against measures proposed by her to improve security and also just happened to contribute a good deal of money to the Democratic Committee. It's the same old tragic story over and over again with politicians and big business the winners- passengers as always the losers. 9/11 didn't have to happen but as is so often the case, greed won out over the safety of passengers, etc. The airlines don't like to increase security measures because it slows the boarding process and they feel they will lose business as a result of that. Of course the irony is that 9/11 happened and that has served to put many airlines out of business and they now have to deal with extra security measures anyway. It is amazing too how the FAA seemed to escape scrutiny considering they were responsible for not passing stricter regulations preceeding 9/11 in their efforts to pander to the airlines.

An additional irony is...many long known aviation safety issues were NOT being addressed prior to 911. Just before this horror struck? progress was actually being made(for public consumption anyway) on some of those long standing issues.

Now? the faa can "realistically" hide behind 911 - in that, they have the "perfect excuse" as to where the attention needs to be focused.

It's a disgrace (to put it mildly) to realize how many times promises have been broken - after an aviation disaster happens. Tombstone Mentality is bad enough.. but when those promises are repeatedly made to deliver "comfort" to those of us who take them at their word and actually hold out the hope they HAVE learned by their past/repeated mistakes?

Reprehensible.

It would seem appropriate on this extremely sad day to bounce this thread up.