# Why does the Boeing 737 have 5 x the passenger fatalities of the Airbus 320?

I thought I'd compare the accident rates of these two similar, competing, very successful aircraft.

                    737     A320

Introduced          1968    1988
Number Built        8104    6171
Accidents/Incidents  329      60
Hull Loss Accidents  154      23
Fatalities          4287     782


These figures are from the respective Wikipedia pages. I assume they have not been the subject of edit-war falsification by fanboys of planemakers.

I wondered what might be the reason for the 737 seeming a much less safe plane.

• The 737 design is 20 years older?

• But it's still in production, so 1988...current 737s can be built with many of the same safety features found in an A320 produced the same year. Were 3,500 of those fatalities in the first 20 years?
• There are lots more older less-safe 737s being flown by corner-cutting budget airlines in third-world kleptocracies?

• Are there that many 40-year old 737s in the hands of incompetent airlines?
• The 8,104 737s have, on average, flown many times more cycles than the 6,171 A320s?

• But they address the same market, why would their usage differ so markedly?
• The 737s have, together, flown many times more passenger-miles than the A320s?

• See above. Plus I imagine older 737s have mostly been retired.

My speculation doesn't seem very convincing to me. Is there a better explanation supported by reputable studies?

• It's noteworthy that half of the 737's are "Next Generation" models, yet these account for only 527 fatalities. The remaining 3760 fatalities were in earlier models. Of course, these planes have flown more miles, so the difference in fatalities per passenger-mile is most likely lower than 5x. – MSalters Aug 27 '14 at 22:45
• and that is a reason to not trust those kind of statistics, every accident/incident/crash must be viewed separately. Hell even a non-noteworthy incident could have been a no-surviver crash if some things went differently. – ratchet freak Aug 27 '14 at 23:00
• Without looking more closely at the numbers, I would say that a major contributing factor is that the 737 contains data starting 20 years prior to the start of the data for the A320. Aviation progressed a long ways in those 20 years by developing safe practices all across the industry which helped to lower the accident rate for everyone. – Lnafziger Aug 28 '14 at 4:01
• worse, it doesn't take into account the average leg length of flights with these aircraft. Many 737s are used for things like island hopping and other very high rotation jobs, and with accidents more likely to happen during takeoff and landing are therefore more "at risk" than aircraft that tend to fly longer hops. – jwenting Aug 28 '14 at 7:46
• "See above. Plus I imagine older 737s have mostly been retired". Take a trip to Central and South America. You'll find plenty of the old 737 in service and being scrapped for parts to keep the other old ones in the air. – casey Aug 28 '14 at 18:14

For one, the early 737s had a design flaw which allowed the rudder to reach a hard-over position, and get stuck there. Boeing initially dragged their feet in rectifying this, so it took a number of accidents before this design flaw was corrected. This is a major contributor to the high number of fatalities in the early operational years.

To have a more meaningful comparison, I would suggest to compare the fatalities per year, so the status in operational procedures can be excluded as a factor. As others have pointed out, this should give a more equal basis for a comparison.

If you look at fatalities per year, beginning in 2000, these numbers emerge:

Now you need to factor in the number of flight hours per type per year, which I do not have available. But I would expect that the low fatality rate in the early 2000s is more due to the small number of A320s around. But if we just compare the 737 NG with the A320, this advantage is reversed, and still the A320 comes out slightly ahead.

But thankfully, there are very few accidents for each type per year, so the statistical base is too low for a meaningful comparison. If we want to draw any conclusions, it could be that the A320 is indeed a safer aircraft, but the certainty of this conclusion is very low, being based on a very small number of accidents.

• Well, to be fair, the difference between the NGs and the A320 here would be well within the margin of error. A single incident would likely flip which one had more fatalities. And, in fact, according to your numbers here, it was the other way around as of the end of 2009. I doubt that which airframe is more safe has really changed since then. – reirab Aug 28 '14 at 14:34
• @reirab: I agree in general; when a single accident can flip the picture, any statistical conclusion would be almost meaningless. But you need to consider that many more A-320 were flying at the time than 737 NGs, and if you take that into consideration, the A-320 could well be considered safer even in the time before 2009 - with all caveats, as mentioned above. – Peter Kämpf Aug 28 '14 at 18:42
• +1 for the last part of the answer. This is a really rare event, and statistic over such low rate are not relevant. – Manu H Jan 26 '15 at 9:25
• Just to further highlight the statistical meaninglessness, the A320 total is now more than twice as high as it was just over 2 years ago, while I think we can all agree that the A320 is no less safe than it was a couple of years ago. The 777 highlights this even more. Until summer 2013, there had been a grand total of zero passenger deaths on a 777 ever (in nearly 20 years of operational history,) then 3 fatal incidents happened within a year (none of which, as far as we know, were actually related to the airframe.) – reirab Apr 9 '15 at 18:30
• We simply need more accidents to collect statistical data. /sarcasm – yo' Mar 21 '16 at 18:21

Raw numbers like this are more-or-less useless. You would have to plot the accidents against the year to even begin to have something reasonable. You also need to remove non-airframe accidents like:

• under-full-control crashes like controlled flight into terrain (pilot screwups have nothing to do with the airframe, and account for a lot of accidents)
• weather-related issues that are not pilot screwups (e.g. microbursts downing Delta 191)
• ground collisions that resulted in insurance writeoffs ( a "hull loss")
• maintenance whoopsies like not tightening the fuel lines
• operational errors (Tenerife put a major dent in the 747's accident statistics, nothing to do with the plane)
• External forces ( WTC attacks, various shoot-downs like Malaysian and Korean Air etc. )

Once you look at 3rd-owner and beyond you may find yourself in the less developed countries with poor maintenance, training etc. and then it gets a lot harder. It's even tricky with first-line carriers. Was Aloha Airlines Flight 243 (the Sunshine Roof flight) a design flaw, maintenance not fixing it properly, operations not grounding it for maintenace etc. etc.

• So, what is the answer? Why is there the discrepancy reported in the question? – user2168 Aug 28 '14 at 2:58
• @Articuno "You can't ever know for certain, there are too many variables to control for." – voretaq7 Aug 28 '14 at 4:03
• Pilot screwups may have something to do with the airframe and so may weather-related issues. The flight envelope protection of A320 does help sometimes. – Jan Hudec Aug 28 '14 at 4:53
• Are you saying that Airbus pilots make fewer screwups, avoid weather, don't hit things on the ground, have more careful groundcrew & engineers, get better ATC instructions and are less favoured by terrorists? I'd expect little or no difference. – RedGrittyBrick Aug 28 '14 at 8:28
• Depending on your objectives, removing weather- or pilot-related accidents might not be the best way to approach these data. Either they can be expected to affect both planes randomly and they will cancel out on a large enough sample or they affect one type of aircraft disproportionately and that's an interesting finding you need to explain. – Relaxed Aug 28 '14 at 10:07

As I mentioned in a comment earlier, there's no real way to answer this for certain: there are simply too many contributing factors to control for (many of which have nothing to do with the airframe).

Broadly speaking I suspect the 20 years more service time, much of it back when aviation accidents were somewhat more common, is a large part of it. Boeing has a great presentation covering the commercial fleet accident rate from 1959-2012, and just looking at the accident rate graph shows how far commercial air travel has come in cutting their accident rate.

That being said while I do enjoy mercilessly mocking Airbus aircraft the A320 series has an enviable safety record, and certainly some part of that can be credited to the airframe and crew training programs in addition to the broader safety improvements in commercial aviation - just don't ask me to be specific as to how much...

• @RedGrittyBrick I knew it existed but Google found it for me (I don't recall my search string, but "Boeing statistical summary of commercial jet airplane accidents" is apparently what it's officially called). A new one should be coming out right about now, and SKYbrary.aero has historical ones which are interesting for the "current" event details. – voretaq7 Aug 28 '14 at 15:49
• The link actually has the answer. Hull loss rate/million departures: 737-1/200: 1.75, 737-3/4/500: 0.52, 737-6/7/8/900: 0.26, A318-A321: 0.26. The old 737s are 7 times less safe. – MSalters Aug 28 '14 at 22:24
• @MSalters Drawing that conclusion ignores the fact that the general accident rate was substantially higher pre-1984 (when the 737-300 hit the market). There are many factors contributing to the decline in accidents, many of them completely unrelated to the airframes in service, and to simply state that the older airframes are "7 times less safe" ignores every other contributing factor. It's like saying the 737 is "5 times less safe" than the A380 based on the raw numbers in the question... – voretaq7 Aug 28 '14 at 22:49

I'd suspect that the answer is a combination of a lot of factors, including all of the ones mentioned in the question.

•There are lots more older less-safe 737s being flown by corner-cutting budget airlines in third-world kleptocracies?

•Are there that many 40-year old 737s in the hands of incompetent airlines?

In short, yes. They might not all be 40 years old, but, yes, there are tons of older 737s that were sold off to third-world airlines and/or airlines that simply operate in more dangerous areas. It's probably the single most common jet airliner in such situations.

•The 8,104 737s have, on average, flown many times more cycles than the 6,171 A320s?

•But they address the same market, why would their usage differ so markedly?

While I'm not sure that their average use cases are very different for aircraft of the same age, there are far more old 737s than old A320s. A lot of the 737 fleet already had flown many cycles before the first A320 ever rolled out of the factory. The average 737 is significantly older than the average A320 and, thus, has flown more flight cycles. While the first A320 delivery was in 1988, the vast majority of A320 deliveries have been since 2000 (5,029 A320s out of 6,171 in service were delivered 2000 or later.) Based on their respective Wikipedia Orders and Deliveries numbers by year, the average fleet age of the 737 is a little over 16 years while the average age of A320s is only about 8.5 years. Of course, this also means that there are a lot more old 737s to be sold to the aforementioned third-world airlines than A320s.

•The 737s have, together, flown many times more passenger-miles than the A320s?

•See above. Plus I imagine older 737s have mostly been retired.

In addition to the average 737 being significantly older (and, thus, having significantly more flight-cycles) than the average A320, you still have to consider that there are 31% more delivered 737s than A320s. In total, the B737 fleet has about 131,000 aircraft-years of history compared to about 52,800 aircraft-years for the A320. So, aircraft-years of service alone explains a factor of 2.5 (half the difference.)

So, in short, I don't think any one factor explains the difference, but the sum of several factors, most of them being related to age, does explain the difference.

A320 has triple redundancy in systems like AOA and radio altimeter/ autoland vs Boeing double radio altimeter/ autoland and single aoa channel. Design philosophy and cost savings approach seems more safety oriented on the airbus.

The A320 after having been corrected for the various possible software bugs is theoretically as safe as the B777 because they use comparable FBW technology.

Technically you cannot compare the A320 to the B737 because they are technically totally different even though competing commercially.

The safety of the A320 relies on the automatic checklist like the B777, on the FBW, on the redundancies as a summary on Technology

The safety of the B737 relies on the experienced pilots from generation to generation who know all its qualities but also all its weaknesses

June 18, 2019

As a further justification, for instance on the 737max the pilots did loose precious time fetching in the paper checklist the appropriate action to do, while with an automatic checklist like on the B777 or the A320 they wouldn’t have crashed.

• Just because an A320 has more advanced technology, that hardly implies you don't need an experienced pilot to fly it. Similarly, the 737 also relies on various technologies and redundancies to improve safety. It is a different concept, but all aircraft require good systems and a good pilot to fly safely. – Bianfable Jun 18 '19 at 11:44
• @Bianfable, what you are stating is the Bible of aviation, and everybody does agree; what I am indicating are features that are more necessary to this or to that aircraft, for instance on the max the pilots did loose time fetching in the paper checklist the appropriate action to do, while with an automatic checklist like on the B777 or the A320 they wouldn’t have crashes – user40476 Jun 18 '19 at 12:02