Here’s a number that should be on the front page of every major newspaper: 224. That’s how many people died–worldwide–in airline crashes last year. Around 3 billion people boarded some 35 million flights, each of them traveling over 500 miles per hour in an aluminum tube 7 miles above the earth. And only 224 died. That’s simply an incredible number.
Some other statistics may put it into perspective:
- Over 400 people died in the United States last year from falling out of bed.
- Over 300 people drown in bathtubs every year.
- About 2,900 people are killed by hippos in the average year.
Looks comfortable, but it’s more dangerous than a Boeing.
This is just the latest evidence that humans are terrible at evaluating risks. Next time you think about taking a bath before getting into bed, consider that you might be safer in seat 34D somewhere over Siberia.
In fact, 2013 was the safest year ever for airlines, surpassing 2012 (which was the safest ever), which surpassed 2011 (which was the safest ever)–you get the picture. The few accidents that do happen are basically flukes, like the seemingly intentional crash of a Mozambique Airlines airplane or the Russian 737 pilot who apparently didn’t have a license.
The business jet record is similarly impressive. There were a total of 6 fatal accidents in US-registered jets (tracking non-US private airplanes is tougher) during 2013, resulting in 17 fatalities. While certainly tragic for those 17 people, this is another ridiculously small number. All of these fatal accidents were in Part 91 operations, so literally no Part 135 jet charter customers died.
The end result is that, if you buy a ticket or book a charter on a jet in the United States, your chance of dying in a crash is almost zero. I may be tempting fate here, but the statistics say that we’ve pretty much won our battle against fatal accidents. That’s not a reason to get complacent or assume pilots are perfect, but it’s time we acknowledge the facts.
That’s something the FAA and NTSB are having a hard time doing. Declaring victory would be, in many ways, a defeat for these organizations in the contest for power and funding. Like a political group who wins their election, it’s hard to move on, and simply maintaining the status quo doesn’t advance a career. The incentive is to do more, to break new ground–which means more regulations, more programs and more cost. But what problem are we solving? We’re far beyond the point of diminishing returns; we’re now in the statistical noise.
So let’s relax on the sleep apnea witch hunt. Let’s relax on the 1500 hour first officer mandate. By definition, the cost/benefit analysis cannot work for these policies (you cannot have fewer than zero accidents). But we’ll all be paying the costs of these–and many other–decisions for years to come.
And it’s not just airline pilots and passengers who pay the cost. Many of these regulations impact general aviation, sometimes in unimagined ways. The sleep apnea policy is just the latest example–an expensive program that fixes a non-problem that is aimed at the airlines, but will end up burdening private pilots and scaring new ones away.
Nobody gets promoted for saying, “Things are great and our work here is done.”
To be fair, general aviation’s accident rate is certainly higher than zero, and it has not been improving nearly as fast as the Part 135 and Part 121 world. There is a huge gap between the safety of that airline seat and the one in your average Cirrus. But I’d argue that the GA accident rate has leveled out for the same reason it has leveled out (at almost zero) for the airlines: we’ve done just about all we can do. All that we can do, that is, without fundamentally changing what aviation looks like in the US.
Want to get that GA accident rate close to zero next year? Mandate two pilots, twin turbine engines, detailed Standard Operating Procedures, rigorous biannual training, IFR flight plans for every leg and Part 25 airplane certification standards. That’s the proven recipe for miraculously safe flying, and it would most likely deliver impressive results for all of us. But obviously it’s completely at odds with what general aviation stands for: we can’t have perfect safety and the freedom of flying a Cub without a radio or a flight plan. It’s not possible.
This is not a defeatist perspective, it’s a realistic one. New technology and better training are great things, and we should always strive to be better pilots. We owe that to our passengers and ourselves. But that responsibility is on us as individual pilots, not on regulators. If you and I manage the risk well, we can fly very safely in general aviation airplanes. We have to recognize that not everyone will, though, and I’m not sure catering to the lowest common denominator does anything other than damage GA. We simply have to accept some risk in exchange for the wonderful freedoms we enjoy as private pilots.
Are the FAA and NTSB willing to accept that? I’m not convinced.
Don’t get me wrong: these organizations should exist and they have an important role to play. The paying public deserves to fly safely and have confidence when they buy a ticket. But there’s a fundamental difference between public transportation and private flying. When we chase shadows in the airline world, there are serious consequences that trickle down to GA. It’s time to be more sensitive to these consequences, and to be more realistic about risk.
- Weather flying means learning to read clouds - November 11, 2024
- What’s wrong with the teardrop pattern entry - September 25, 2024
- What matters for VFR proficiency: better landings - August 26, 2024
Very well done John. I’m not sure anyone could state this better.
Frankly, what the FAA wants is all GenAv pilots be IFR certified. The “Sport Aviation” license would be for VFR flying and continue to be as restrictive as it is.
Agree with your points. It seems there are many instances right now in America where ‘personal freedom’ and ‘personal responsibility’ are battling government intrusion. Healthcare, guns, education, energy production, and many forms of transportation come to mind. To say the least, it is a critical time for America in terms of what She stands for.
I go back to the discussion we’ve had many times in AirFacts regarding the massive regulation differences between driving a 5,000 pound car or SUV and flying a 3,000 pound airplane. Why is one virtually regulation-free while the other requires detailed regulations for both machine and operator? Is it because less than 1% of us are pilots, and society as a whole is fearful of flying? No one has yet satisfactorily explained to me why there is such a stark difference in regulations between driving and flying. Can anyone help me out?
It’s a great point John – I suspect several things are at play here – the most obvious is ‘level of difficulty’. We can give a 16 y/o a couple hours of instruction and they can safely get in a car, drive to the mall and return home likely unscathed (let’s ignore teenage accident rates for now). But we can not do that same thing in an airplane. Flying aircraft requires a much higher degree of training that driving a car.
You said it yourself–bureaucrats being bureaucrats !! All those lawyers on the public payroll have to do something to justify their existence.
Well – I totally agree with the overall safety of flying…but these numbers are likely very misleading – and certainly out of context…
“Over 300 people drown in bathtubs every year” – that’s only relevant if we now how may baths were taken in the year. I’m guessing most everyone that flew – took a bath…
Absolute numbers aren’t all that useful when managing risk.
okay – I get I’m nit picking – the crux of the article is – yes aviation is safe..but there is better data to support that fact.
You make a fair point, Gary, and I’m obviously using those number for illustrative effect. But 3 billion people flew on airlines last year, so it’s not exactly a small sample size. Put another way, out of the 35 million commercial flights last year, 99.9993% were completed without an accident. I’m not sure I tie my shoes correctly 99.9993% of the time!
your absolutely right! I love the positive spin of that sentence…”out of the 35 million commercial flights last year, 99.9993% were completed without an accident”
“Over 300 people drown in bathtubs every year” – that’s only relevant if we now how may baths were taken in the year. I’m guessing most everyone that flew – took a bath…”
Not only did they take a bath, it’s more likely a few people have taken more one each.
It’s difficult to drown in a shower, so I’m not sure those should count. I have no idea how prevalent bath-taking is as opposed to showers; but I know I’ve flown about 20 times as often as I’ve taken baths in the last 25 years.
I’m keeping my eye out for hippos.
The perverse irony with aviation safety is that as aviation fatalities become less common they become more remarkable.
We all know that aircraft accidents, even non-fatal off-airport landings that damage little more than egos (if that) generate a lot more media attention than the equivalent in ground travel of the ubiquitous “fender benders”. A forced landing by a private aircraft on a public highway often makes the national news. The impression left by the media is that GA aircraft are falling out of the skies quite regularly, in part because they do so quite irregularly.
GA safety has nevertheless improved substantially in the last two decades – per NTSB data, the GA fatal accident rate per 100,000 hours flown has dropped from 1.81 in 1994 to 1.24 in 2011 – a 31% reduction, which ain’t nothin.
On the other hand, USA auto fatality rates per 100 million miles per NHTSA have also dropped a bit more over the same timeframe … from about 1.73 to 1.10, for a 36% decrease. The improvements in safety results in both modes of travel over this era seem fairly comparable.
Obviously there is still a lot of room for improvement in aviation safety … just as there is room for improvement in driving safety. So why has the personal travel safety record been improving? A few factors seem obvious:
New cars have become much more crash resistant in the last two decades, while air bags, anti-lock brakes, traction controls, and so forth have done quite a bit to reduce fatalities in auto accidents. With GA, the arrival and popularization of in-cockpit GPS navigators, real time weather information, glass panels, TAWS and TCAS (even as delivered by the portable units that are more common in owner-flown planes) and such have certainly helped reduce certain kinds of aviation accidents.
“Fixing stupid”, however, as Ron White likes to say, is much harder. And that is why neither automobile nor general aviation accidents will ever approach zero, as they have with professional airline crews.
So why do the airlines do so much better than GA in avoiding the avoidable accident? They do it with much more stringent FAA rules, professional two-pilot crews with extensive type-specific recurrent training, company SOPs, and much wider adoption of improved cockpit technology. Even so, the pros are still capable of occasional bouts of stupidity (re: Asiana 214, Air France 447, etc.), though the potential for such stupidity is mitigated considerably.
Declaring victory doesn’t mean their work is done. They don’t have much ground to be gained, but there’s plenty of work in maintaining the accomplishments that have been made. There’s infrastructure to be maintained, there are ongoing operations to be policed to ensure that they maintain the standards that have been set. If the regulators want to focus on something it should be fighting regulatory capture (good for the public and the rank and file, bad for the top execs if it costs them million dollar jobs in the private sector), and making sure the existing regime is kept as airtight as possible.
I don’t know…I think the 224 people might disagree with you.
I don’t know…I think the 224 people might disagree with you.
“This is just the latest evidence that humans are terrible at evaluating risks. Next time you think about taking a bath before getting into bed, consider that you might be safer in seat 34D somewhere over Siberia.”
The author must have been talking about themselves when they say humans are terrible at evaluating risks. 350 million people in the US, 365 sleeps/year, 400 fatalities, means 1 death per 319 million sleeps. 3 billion plane rides and 224 deaths means 13 million plane rides per death. For US citizens this means sleeping is still safer than riding a plane by a factor of 24.
Andrew, as I mentioned above, those statistics are used to make us think–they obviously wouldn’t hold up very well in a Statistics 101 class.
But I still say we are terrible at evaluating risk, and I think there are mountains of evidence to back it up. Again, 99.9993% of all airline flights landed safely last year; among US and European airlines, it was 100%. Yet thousands of people are paralyzed by fear of flying while everyone sleeps soundly in their beds.
At the very least, it’s worth celebrating how safe we have made flying on the airlines. In less than 100 years, it’s gone from dare devil to incredibly safe.
John, I don’t think you get the point. If it wouldn’t hold up in a Stats course then that discredits your “statistics”.
Have you heard of the book “How to Lie with Statistics”? Your article here is an excellent example of how to lie with statistics. You should read that book.
This is why we pay people to study statistics… so that we get accurate statistics. You don’t have a strong point if all your stats aren’t compared properly.
OK, let’s try this: the fatal accident rate for airlines in the US and Europe is 0%. Literally zero. Does the average American know that? I’d bet a lot of money they don’t. We are terrible at judging risk.
Again, I’m not saying we shut down the NTSB or anything, but we need to consider what we’ve accomplished. There is a formula for incredibly safe flying. It’s proven.
The question (to me) is: do we want that formula applied to GA or are we willing to accept some increased risk in exchange for increased freedom?
The accident rate between accidents is always 0%. That’s slightly tongue-in-cheek but it’s true… you can’t get any data if you assess a period of time where there are zero accidents. The longer the time since the last fatal accident the better, however, you MUST include time in there or the statistic is meaningless.
I do agree that the airlines are very safe these days. But GA is not even close to being that safe.
The stats on bath tub and sleeping accidents aren’t even close to relevant. Almost stupid to mention them. Most of us spend way more time in bed or a bath tub than we do a plane.
However, agreed the 121/135 safety record is impressive, but GA part 91 is still risky. However, I’ll take the risk with it’s freedoms anytime over regulation. But doesn’t mean we still can’t strive for better safety.
And a lot of the “airline” stuff we can have. We can have an ops manual, good training, second in command (for us that fly with a mate), and our ops manual can give us safety parameters to live with, with little unnecessary restrictions. But… most of us don’t do these things.
I won’t consider flying safe until all planes have giant spider silk parachutes, in case they fall out of the sky.
The problem with a fear of flying is that, unless it’s mild, the irrational mind locks on to ANY fatalities and believes they WILL be the next one to perish.
However, this information will help people with flight anxiety or mild fear of flying, so many thanks for sharing it.
Neil Shearing
Safe, in terms of the number of crashes per 1000 flights, but it has risks and it’s not sustainable. What do I mean by risk? What is the survival rate of the victim of a plane crash relative to someone falling off a bed? Airline travel is bad for the economy. I would take a flight from the east coast to the west coast, maybe, but for regional travel I would prefer high-speed rail. People in Pittsburgh should be able to work in Chicago and people in Detroit should be able to work in Philadelphia, and people in Philadelphia should be able to commute to New York. Why don’t we have an advanced rail system like Europe or Japan?
“Why don’t we have an advanced rail system like Europe or Japan?”
Lots of reasons, but one huge factor is that the USA is a vast place; Europe and especially Japan, by contrast, are small places.
High-speed rail also fails at the intersection of physics and politics (a.k.a., fantasy). Example: “people in Detroit should be able to work in Philadelphia.” Okay, let’s do some arithmetic… Distance – about 585 statute miles. Presumed maximum allowable daily one-way commute time – 90 minutes. Required average speed – 390 mph. But here’s where politics enters the calculations: there will be no express train from Detroit to Philly, because the politicians in the in-between communities (like Toledo, Cleveland, Pittsburg, and Harrisburg – to cite just the most-obvious / most-worthy examples) will insist that the train has to stop in their precious hamlets. The resulting extra 40 minutes of wait-time (to say nothing of accelerate/decelerate time allowances) would boost the required average speed to 702 mph.
Seen many Mach-1 trains cruising around lately? No? Well, that’s why it’s complete fantasy to postulate that “people in Detroit should be able to work in Philadelphia.” And I haven’t even touched on the required logistics for the train stations….
In the very early days flying was inherently dangerous. It’s only safe today because of the excruciating focus put on safety on all levels. Without the massive safety framework and current attitudes, I would wager that flight safety would deteriorate very fast. In any case, with the ever increasing commercial air traffic, new systems and policies are more important than ever. Flight safety should never be taken for granted.
I agree – flying didn’t become safe by chance. But there’s a difference between staying vigilant and constantly introducing new regulations and inspections.
So, we’ve gotten our arms around the technical end of GA safety. That leaves the human to be fixed. AARRGGHH!!
Somehow, we need a gauge (steam or glass) that indicates how foolishly risky a person is on any given day, and over a protracted period. Then, some range markers for green, yellow, and red. Insurance rates commensurate with range marker color would help motivate some folks. Etc.
Life is risky; if we want to participate, we can expect a skinned knee now and then. And, for some of us, the finish line is closer than for others. We cannot legislate Life – but that doesn’t stop bureaucrats from trying.
TOTALLY MISSING: ANY MENTION OR SUGGESTIONS FOR IMPROVING OUR CONTINUING EXTREMELY HIGH RATE OF GA FATALITIES. (WE ARE UP THERE WITH MOTORCYCLES !)
If GA pilots would stop buzzing their buddy’s homes, quit continuing VFR flight into IMC, and land before they ran out of gas, a huge proportion of GA accidents would go away. Something on the order of 80%, maybe more. How to do that, without imposing regulations that would make GA essentially different from what it is today, I do not know.
There are already regulations in place that tell us not to do any of those things, but we continue to do them What would the technological answers be? Ground control that monitors fuel state, ground speed and distance to go so that planes can be ordered to land when low on gas? Give FSS dispatching authority over every GA flight? Put in “envelope protection” that won’t let you descend below 1,000 ft. AGL unless you’re within 5 miles of an airport unless you’re transponder is on 7700? Honestly, for lack of a better term, I have to use the old hackneyed question: How do you fix stupid?
How about starting with ramp checking each pilot who shows up?
That should reduce a significant portion. It would remove those without licenses / current medicals / no bi-annual / not current. Laughing now ? Well think, how many do you know or think are flying without proper credentials ?
By the way , I mean all pilots and also check to see if they are rated for what they are flying.
Laughing now ? How about the recent airline accident where the pilot was not rated ?
Like I said, this would make GA essentially different than what it is today. It would require a whole new infrastructure worse than the TSA because it would mean having every airport staffed 24/7, wouldn’t it? Or would the pilot(s) have to phone in to the FSS, give their details, receive a dispatch code and give that to ATC before taking off from a towered airport or departure after takeoff, making flight following mandatory. This kind of thing is what I meant by “dispatch authority”.
It would be a horror in many ways. Are we going to expect the feds to know every type of aircraft that qualifies as an LSA and therefore not require a medical? What about homebuilts that meet the LSA rule?
Personally, I think accidents caused by people who have no license, are out of currency, don’t have a medical, etc. are fairly small. I think that the majority of the pilot population is not suicidal and won’t fly when they’re truly incapable of keeping the aircraft and themselves in one piece.
David:
What you are discussing is the norm in many, if not most countries now. In the U.S. we are used to “Free Flight” and VFR.
The Feds don’t like un-controlled flights. Remember, more people, more power and a bigger department budget, meaning more money more power !!! This is a simple fact, practiced by both governments and private industry.
Everyone must realize, the system is slowly being redesigned to allow only IFR flights.
I wish I could agree with your last paragraph, unfortunately many accidents support the reverse. Kick the tires and light the fires, oh, everything will be fine, I only had 4 beers. I take my heart medicine every day and haven’t had any symptoms in the last year. I test my sugar level all the time and rarely have any problems. Sound anything like the last 100 or so accident conclusions ?
If cheaply airlines offered half price tickets for twice that risk, I’d be taking the family to Disney land.
These are all excellent points. If I might throw a couple more into the mix….First a year is probably too short a sample Tenerife was the first fatal KLM had had in 19 years (mind you it has not had a fatal accident in the 36 years since that tragic day). But as others have already implied before “there are lies, damned lies and statistics”.
Yes the accident rate is excellent – and I plan to use the hippo analogy every chance I get to point out the safety of aviation but you also have to recognize that although in the case of airline ops in the “zero hero” regions globally the the picture is less impressive.
So two questions are posed how do we prevent complacency from creeping in (what I call “we’re number 1 why try harder syndrome”) and second bring the airline industry as a whole to the sunlit uplands that are a zero accident rate.
I’d argue the former is about changing the accident reduction process from the reactive to proactive – the former has served aviation well but has probably run it’s course. In the case of the latter the picture is more challenging and must recognize a number of factors that may not be palatable and are certainly contentious yet nevertheless they must be dealt with if the zero rate is to be reached.
I add an important caveat at this juncture (just so we are clear!) – I firmly believe that safety improvement should be paramount. But safety improvement (or safety maintenance – might be a better description) Has a cost. Problem is the law of diminishing returns comes into play, as safety rates have improved so the ROI on that investment suffers.
For more years than I can remember advocates of safety at pilot groups (and I was one of them) used the mantra “you think safety is expensive try and accident” to underpin our arguments for this program or that one. We even went so far as to define what that cost might be. Naturally by definition the data was ballpark in nature it is doubtless an exercise that has also been carried out by most airlines. My point being eventually (if they haven’t already) somebody is going to ask – if our exposure is Xbn why spend 1.5X bn to prevent something that might not happen?
It’s a tough sell “how do you improve on zero” – unless you are talking about band aid measures like the 1500 hour rule (square root of zero use in my view) it is unlikely that more effective measures will be sanctioned (and funded).
OK back to GA and the GA accident rate – I have a question – if it means giving up more freedoms (though it might not feel like it those of you in the US enjoy more freedom to fly that just about anywhere else) is there really a benefit to the aviation fraternity and the wider community that would make such a sacrifice worthwhile? I submit the answer is no and here’s why: The stupid will do stupid things with unfortunate results no matter what regulations are in place while the prudent pilot suffers under the yoke of ever more stringent measures.
There never will be a “0” accident rate—aint-a-gonna happen.
On the bright side, it gets rid of a lot of stupid.
Incidents are just lucky outcomes–like Southwest landing on School of the Ozarks mountain–just dumb luck they made it.
Stupid has been with us forever and will continue. Like speeding and following too close on snow or jumping off bridges.
Is there any evidence that accident rates are lower in countries where VFR flight like we know it doesn’t exist? Possibly that’s true, but I’d bet the majority of flights are being carried out by commercial crews, not individuals.
And just because that’s the way “the rest of the world” does it doesn’t mean we should roll over and let our government follow suit.
I still maintain that if we as pilots just stopped doing low level maneuvering flight, didn’t continue VFR into IMC conditions, and stopped running perfectly good airplanes out of gas, we’d reduce the GA accident rate to a fraction of what it is now. As I stated earlier, there are already regs against doing all those things now, so what changes would solve those issues?
For those who don’t intend to be stupid, the best safety investment is more avgas. When I evaluate improvements in my aircraft I convert the cost into flight hours not flown. It is a rare item that beats the flight hour investment. Now if we could only get those trying to “help” to understand the same concept.
Comparison with bed fall deaths is incorrect because it lacks common denominator – which is time spent in bed / aircraft. What’s higher – average number of fatalities per total human-hours spent in bed or total human-hours spent in aircraft? Let’s compare. Let’s assume people sleep 8 hours a day on average and average flight duration is 10 hours (I know it’s a stretch but it only reinforces the point). So here is your chance to die by falling out of bed per hour of sleep: 400 / (8 h x 365 days x US population size i.e. 323 000 000). And here’s your chance of dying per hour spent in aircraft: 224 / (3 000 000 000 people x 10 hours).
So even with this “average 10h flight” stretch sleeping in a bed is 17.6 times more safe than sitting in a flying aluminum sausage so stop feeding us yellow snow please.
excelent article