Ok, I’ve heard for many years that flying commericial is much safer than driving the same distance.
The reasons for this are intuitively obvious. If you drive, even if you are the best driver on the planet, you are in close proximity to high speed metal boxes of death driven by anyone who has enough money for gas and passed a basic test given decades ago.
At any time, especially on those highways divided by just a painted line, someone could fall asleep or drop their phone or do any number of stupid things and plow into you head on.
If you fly, a professional who usually has a lot of experience and training controls the vehicle. The pilot also has racks of expensive computers and instruments to help him perform the tasks. Once in the sky, there’s countless cubic kilometers of empty space such that collisions are very unlikely even without air traffic control.
But…what. Let’s suppose you need to travel 1000 miles to reach your destination. According to this, your chance of getting killed is actually slightly higher. (I’m looking at fatalities per 100 million miles)
What gives? Where are the stats that say otherwise coming from? I read a studythat fear of flying after 9/11 causes several thousand additional traffic fatalities.
Admittedly, the numbers are close enough that I would rather snooze on my headrest than deal with traffic, but in absolute risk it does not appear to be the safer choice.
You have to be careful what you’re comparing and what assumptions you are making.
First, you need to define “dangerous.” Is the only danger from travel death? Not at all. Even so your first link uses fatalities as the only indicator. Injuries from automobile traffic run nearly 100 times[!] higher than deaths. Shouldn’t that be considered dangerous? In commercial airline crashes, the number of injuries does not exceed deaths by two orders of magnitude.
Specifying commercial airline crashes is another critical point. It is well known that commercial air traffic is remarkably safe. As your cite notes, many years the number of fatalities is 0. The larger number of deaths come from General Aviation, all the small planes, with statistically less well-trained pilots and less well-maintained aircraft, often flying without air traffic control into major properly lighted and market airports. They separate out those numbers.
And that’s using 2005 numbers. The rate of both injuries and deaths from cars has plunged in more recent years according to the link above.
Deaths from cars are probably now below airlines per 100 million miles. The *danger *from cars, however, remains far higher.
Your first link is about General Aviation (i.e. including small, single-engine planes and private pilots flying themselves) - not commercial passenger service (mainly multi-engine jet airliners, piloted by two commercial pilots, etc).
The challenge here is figuring out what’s a fair comparison.
If you compare taking an airplane from L.A. to Hawaii with driving your car to Hawaii, obviously that’s stupid because you can’t actually drive from L.A. to Hawaii. But what about going from L.A. to N.Y.? Suppose you live in L.A. and you have family in N.Y. and they invite you to come for a visit. How often do you go, and how often do you decide to stay home? If you have access to commercial airline flights, you might go to N.Y. twice a year but if you’re forced to drive the whole way you might only go twice per decade. So, if that’s the case, then what we should be comparing is what’s your chance of injury/death from two car trips vs. twenty airline flights of equal distance.
Also, the death rates for automobiles usually lump into one big pile all the deaths from city driving and highway driving, but really they are two different rates. I’ll bet city driving is much more dangerous. When you’re comparing LA-NY trips, nearly all of it would be highway driving.
So let’s use some numbers which I’m just gonna make up…
city driving 4 deaths per 100 million passenger miles
highway driving 1.5 deaths per 100 million passenger miles
airline travel 1 death per 100 million passenger miles
So if you fly LA to NY 20 times (120,000 miles) you have a .00120 chance of dying, but if you drive LA to NY 2 times (12,000 miles on the highway) you have a .00018 chance of dying, so airline travel is about 7 times more dangerous.
Again, these are numbers I just made up to illustrate the point.
Those statistics are just for deaths. How many car accidents have left people paralyzed, etc., yet still alive? Any argument regarding safety has to take that into consideration too.
sbunny’s logic above that you should compare 20 air trips to 2 driving trips is not logical in any sense. It’s akin to this (clearly bogus) logic:
Assume you’ve got $1000 to spend on travel. You can afford to either take 5 cheapo Southwest flights or one expensive legacy airline flight in first class to the same destination. Clearly the risk is roughly 5x higher on the 5 Southwest flights. Therefore Southwest is 5x more dangerous than the legacy airline. Clearly that’s nonsense. SWA is incrementally more or less risky per flight or per mile than any other mainstream US airline, but
trying to mix dollars in there is a huge red herring.
But he/she is not too far from making a valid point. Which is …
A good argument can be made that air travel risk should not be measured on miles traveled, but rather on flights taken. The two justifications are that 1) almost all risk accrues to the landing or takeoff phase, and that risk is independent of the cruise duration or mileage, and 2) a flight is a unit; you can’t be exposed to just the risk of one part.
And to answer the OP … no, by no logic can you demonstrate that US major airline air travel is less safe than driving. Driving loses by a large margin no matter which way you validly slice the data.
You just made an argument from authority you don’t have. (airline piloting doesn’t make you an actuary who is the relevant authority here).
Actually, I just link a reference to a way to slice the data in a way that causes driving to win. Let’s make it apples to apples :
You need or want to travel X miles in your lifetime. All of those miles, you can drive or fly. From the numbers I gave, your absolute risk of dying is slightly higher if you fly.
That’s not a “large margin”. So are you claiming this site sliced the data wrong? That’s very possible.
Again, general aviation–which includes all those guys in tiny private planes–vs commercial aviation.
If you just compare “flying” vs “driving”, then yes, flying is more dangerous. It is more dangerous to fly in a dinky two seater aircraft piloted by your buddy than it is to drive. That is clearly true.
But if you compare commercial aviation to driving, then the numbers change. A lot.
The scientific American study must be using different data. It concludes that “flying is safer than driving for trips greater than 18 kilometers…”
The reason it’s wrong is that 100% of commercial airline flights are for trips larger than 18 kilometers. So out of all these flights that are longer, covering a large number of miles, the number of people killed per mile traveled is higher.
The only way the Sci Am conclusion could be true is if there were a huge number of commercial airline flights less than 18 km driving up the death rate. Since there are essentially 0 of those (I can think of no business reason a commercial airliner would ever offer flights for a distance that short), this isn’t true.
This is logically inconsistent with Sci Am’s conclusions. They must have been using different raw data.
I am comparing commercial aviation. The link in the OP analyzes general, but this thread is not. This discussion is about commercial aviation versus driving, anything else is off topic.
Assuming you mean the American Scientist (not SciAm) article quoted in this post - that link no longer works, but the claim in the quote is perfectly reasonable. They are simply saying all the data points for airline are lower than automobiles, but if you extrapolate that line down to 18 km, they do intersect.
You’re still comparing the wrong numbers. Everyone is comparing the wrong numbers. The number of people killed is entirely irrelevant to your personal safety, it absolutely should not be the basis of discussion simply because of the wildly divergent number of people involved in the two types of accident.
What affects your peronal safety is the rate at which the vehicle is involved in an accident which is fatal to you. Plane crashes can kill hundreds of people at a time, but you’re no more or less dead than if you’re involved in a single person car crash that led to your singular fatality. The only number from which we can draw any meaningful conclusion would be the result of dividing the rate of fatal accidents per distance traveled in a vehicle by the probability of you dying in a fatal accident which has occurred to your vehicle. I’ll admit I neither have that second number nor know where to get it, but we can’t arrive at it from what has been posted and I hope its importance is obvious.
Because the number of people killed per mile is what society cares about, its how safe the population of travellers is, but we only need to make plans based on what is more or less likely to kill ourselves and our families, as long as we aren’t dead the deaths of other people on other planes and other cars is a collectively solved problem and not an individually solved one.
That’s exactly what these numbers are: fatalities per passenger per million miles traveled. (Even though they’re often sloppily labeled as just “fatalities per million miles”.) It’s the number of fatalities divided by the total distance traveled by those vehicles AND the average number of passengers in those vehicles. This represents exactly what you’re asking for: the probability that YOU will be killed if YOU travel a million miles using that mode of transport.
That’s not true. There’s a Wikipedia list of shortest flights which has 6 commercial air routes less than 18 km long, with the shortest just 2.73 km. (These short fights generally involve travelling between islands that aren’t connected by bridges, so there’s no alternative route by road.)
As has been pointed out upthread, it is hard to create fair comparisons. One of the best ways to do it is to clarify what decision you are thinking of making. Are you trying to make a personal choice which way to travel between two cities? You need to compare scheduled commercial passenger flights to highway road accident statistics. Ideally you want a combination of passenger miles and per-trip to take into account the increased risk taking off and landing. (In most countries long-distances coaches actually pip commercial air flights, which are about the same as rail, all of which come out far ahead of private automobiles).
Are you trying to decide where government can get the best return on investment to increase public safety? You probably want the fatalities per year. (Yes, there is another side to the equation, but raw risk is a good proxy for cost/benefit of improvement).
Are you just looking for a safe recreation activity? You want the fatalities per flight-hour or driving-hour.
Just remember that aviation statistics are very different:
for general aviation versus scheduled commercial passenger flights
for world-wide versus individual countries
And that the noise in the statistics is huge, making comparisons between airlines meaningless.
