Menu Sign In Contact FAQ
Welcome to our forums

Sad but quite interesting (French accident statistics)

LeSving wrote:

On average, each year 1.7 person per million citizen in Europe dies in a GA accident. This is nothing compared with just about any other activity you can possibly do.

1. It’s absolutely true that people who do not fly in GA (that is: almost all people in Europe) have a very close to zero risk of dying in a GA accident. Hint: If you divide the number of deaths by square meter the number will be even smaller
2. I completely disagree with the second part of your statement: I’d bet there are much less people in Europe each year dying in an accident while driving a 86 corvette or while fly fishing with a 9’ Orvis rod. (Yes, if you think the comparisons are “unfair” this is exactly my point. You always need to have a baseline that accounts for frequency of the activity – again: if few people do something and here only few people die it doesn’t say this activity is safe).

If you are not yet convinced: According to the metrics you use (deaths per citizen per year), Russian Roulette is a much much safer activity than flying GA planes.


I cant see how that population table is relavent. I thought that the accepted standard was fatalities per 10000 or 100000 flying hours ?

rygb wrote:

How do you know that there aren’t two different groups here: one group that is prone to risky flying and stupid decision making … and a separate group that is less prone to those things due to being more thoughtful

There’s simply no evidence for existence of these two groups (also see below): Reading accident reports and looking at all of the (few) pilots I knew personally that were killed in LOC or CFIT accidents, I would not have the impression, that these pilots wer particularly risk taking or stupid. They are quite average pilots – sometimes even described as quite cautious.
Might be, that there is a quite small group of pilots that are extremely stupid and/or risk taking – but these are not the ones driving the accident stats.Vast majority of LOC and CFIT accidents are caused by completely normal pilots who have a bad day or take a few very bad decisions…

Having said that, there are obviously some groups more prone to accidents – and some of these can be connected to risky flying and stupid decisions: E.g. the German microlight community has a very broad culture of ignoring safety relevant regulations like MTOW – not surprisingly the accident rates are significantly higher than in the non microlight segments.


@malibuflyer please check your PMs, or whether your email address on EuroGA is valid.

Shoreham EGKA, United Kingdom

Malibuflyer wrote:

If you are not yet convinced: According to the metrics you use (deaths per citizen per year), Russian Roulette is a much much safer activity than flying GA planes.

I’ don’t disagree with that. As I wrote, it’s a fruitless discussion without proper data. So, using the data we have, we can only draw the conclusions the data permits. Besides, it’s one fifth of 1.7 as Juju points out, or about 0.34 fatalities per million citizen per year. One death every third year per million citizen. That is what we have. As a recreational activity, that the entire society in principle can participate in, and we have one fatality every third year per million citizen. That is hardly anything to write home about as a problem the society should be concerned about. Cancer, car accidents, crime etc those are the killers by orders of magnitude.

GA is a choice, a personal choice. It’s not something that hits you indiscriminately (like cancer for instance). We know it is dangerous, but we don’t know how dangerous, not exactly, although we have some idea. It’s about as dangerous as driving a motorcycle, or boating in the open sea, or hiking in the mountains. It’s comparable to those things in many ways, even though the exact danger may be very different, as well as the total hours of exposure per year. It’s the risk per year, per person that counts in the end.

To survive, whether it is in the air, in the mountains or at sea, is also very much a personal thing. We know what it takes. It requires training, currency, the right equipment for the mission and most of all airmanship. That’s all there is to it, and no statistics will change it. If you do what it takes to survive, you stand a great chance of doing just that. If you don’t, you stand a large chance of becoming a part of the statistics. That is the main thing statistics will show, regardless of flight hours or pilots in a population. The numbers are too small anyway to really get anything significant studying the details.

Just like participating in GA is a personal choice, surviving GA is also a personal thing. But surviving is not just a choice, it’s also very much a part of the community. You won’t survive if you don’t know how, and you won’t survive if that isn’t prioritized enough. This is about the community, the culture. For some, living an exciting (and free) life is more important than worrying about not being alive. Small cultural differences will make large differences in the statistics. This is not really about taking risk for the sake of taking a risk, as it is about not worrying too much about the consequences of that risk. Everybody will die, and there is nothing we can do about it. Far from everybody lives happy lives, but that we can do something about, even if it kills us.

We cannot really change the cultural aspects in this. What we can change is that people die simply because they don’t know how to survive. Better training, better education in everything airmanship and so on.

Last Edited by LeSving at 21 Dec 11:57
The elephant is the circulation

Malibuflyer wrote:

There’s simply no evidence for existence of these two groups…

I’m not really trying to claim that there is – as you say, there is no evidence. I’m just questioning whether drawing certain conclusions from the statistics is valid by pointing out some assumptions that might be being made from the data for which there is no evidence.

What I’m saying is that in the lack of evidence that pilots are all the same (ie. that the relative risk of different factors apply equally to all pilots), I don’t think it’s valid to dismiss addressing the less common categories just because the more common categories are more common, since that is only known (from the data) to apply in aggregate and not to individual pilots (this the data cannot tell us). Further, the cost to reduce the risk of the lower categories may be less than the higher ones – so it may even be that more lives can be saved by focusing a particular spend on the lower categories than the higher ones.

To reiterate, I’m not claiming that my hypotheses above are true. I have no evidence for that. I accept your anecdotal evidence, but as far as I understand we have no hard data telling us what would be the most effective countermeasure. My point is that the data must not be used to infer that addressing midairs and engine failures can be a low priority, because it cannot tell us either way.

Last Edited by rygb at 21 Dec 14:58
United Kingdom

One could complicate this question indefinitely and in theory there are many tiny things to be considered. In practice, however, things are quite simple if one looks at the data:

According to the Nall reports* LOA and CFIT are by order of magnitude (as in more than ten times) more likely to kill a pilot than Midair or non pilot induced (like fuel starvation) engine failure. Let’s for simplicity assume that for every pilot that dies from MA or EF there are 10 pilots who die from LOA or CFIT.

Now in theory you are completely right that it might be possible, that this one death from MA/EF is still easier (i.e. less costly) to avoid than one of the ten deaths by LOA/CFIT. But how realistic is that in practice?
And in theory it is of course also possible, that out of those 10 LOA/CFIT deaths, 9 are from risk taking jerks and idiots, so that for the deliberate pilot the risk distribution is more 1:1. Again: How realistic is that in practice?

*the best source for GA accident data and while strictly only applicable to the US there’s little reason to believe that causes of accidents in EU ore elsewhere in the world are completely different


Malibuflyer wrote:

Now in theory you are completely right that it might be possible, that this one death from MA/EF is still easier (i.e. less costly) to avoid than one of the ten deaths by LOA/CFIT. But how realistic is that in practice?

I think you’ve asked exactly the right question. I think it’s perhaps more realistic than you think. We’ve been focusing on LOC/CFIT for a long time. As you say some pilots you knew did not seem the risky type. Perhaps the cause of the majority of “careful pilot” LOC/CFIT is that some or all pilots have a tendency to temporary lapse in judgement or control ability that is extremely difficult to train out. Perhaps we’re already doing all that is feasible. On the other hand, one path to reducing midairs is obvious and “easy”: mandate EC.

If both of these hypotheses are correct, I don’t think it’s unrealistic to consider that addressing midairs might be cheaper per life saved than LOC/CFIT.

United Kingdom

rygb wrote:

On the other hand, one path to reducing midairs is obvious and “easy”: mandate EC.

Unfortunately this is not at all “obvious”: As a matter of fact, the ADS-B mandate in the US had no measurable effect on midair deaths so far – might be too early to tell and there is a small effect but it’s definitely not the end of all midairs. Even TCAS-II equipped planes have midairs unfortunately…

Reason for smallish effect of electronic traffic gadgets in GA (beyond e.g. the great things flam did for gliders soaring under the same cloud) in my opinion is threefold:
- Training of GA-pilots is insufficient. Reading and reacting correctly to traffic warnings is not a simple task. One needs to have a standard collision avoidance drill as soon as the box sounds an alarm (like commercial pilots have with TCAS). If you look out of the window after the alarm, odds are high that you crash before you identify the traffic visually – and that’s what most pilots would do.
- Most GA midairs happen in phases of flight, where traffic systems are less helpful. When I’m in the pattern of a crowded airfield, I know that other planes are around me without a blinking box….
- There are downsides of these boxes (connected to #1): Many pilots spend far to much of their valuable time focussing on this box or trying to visually identify planes which are not a factor at all (but as they are on the screen we feel the urge to see them in the air as well).

Don’t get me wrong: I love my ADS-B and TAS – but I would not expect a statistically significant reduction of risk from that.


As the OP in this thread I have of course followed every twist and turn of what can and should be done and as I have time on my hands due to being unable to fly at the moment following a slight accident which has left me temporarily( I hope) with loss of sight in one eye. The result not of aviation but of DYI. I probably do about the same number of hours of DYI as I do in aviation so the limited statistics in my case point to aviation being safer than DYI.
But sticking closer to the thread, although the figures in the dgac report do break out the numbers of fatalities by type of a/c (ie plane,ULM, helicopter, balloon, glider etc) it is not possible to relate it to the number of hours flown as in ULM flight for instance it is not necessary (in France) to record hours flown in a log book or anywhere else. There are a number of other areas where ULM flying in France is much less regulated and restrictive than in other European countries. Could the price of this freedom be behind what appears to be a much higher rate of fatal accidents in France than in those other countries, or could it be simply that France has more pilots. (Sources: FFA, FFPULM, DGAC.)
The French statistics mirror those of the USA to some extent eg Alaska has more aviation related fatalities per annum than France. The number of hours flown in light aircraft in Alaska is very high as it is considered the major form of transport in this part of the world ( Source: Aircrash Alaska). One might therefore assume that many of the pilots in these accidents would have high levels of currency. The other interesting statistic is that the majority of those fatal accidents in Alaska are due to “loss of control in Flight” (source:NTSB ). They also point out that when they say loss of control in flight, they usually are referring to a stall.
If lack of currency were a problem how many of the contributors to this thread would agree to there being a minimum number of hours regulated at one per month or 12 per annum and what might the effects on GA as a whole in Europe?

Sign in to add your message

Back to Top