Like others who commented on that, I am always bemused by people who fling utterly absurd "statistics" into a discussion without a second's thought along the lines of "hang on, that can't be right". That was my initial reaction to seeing the assertion. It took googull about 1/10 of a second to bring up a
table of EASA and UK National licence holders broken down by age and sex (sorry, couldn't resist) and another 30 seconds to work out that if the assertion were true, we would be reading about an average of somewhere between 2 and 3 PPL/NPPL etc holders dying while flying every week, assuming that nearly all PPL holders hold both UK national and EASA licences (see the note at the bottom of the table.)
If 3 ppl holders died every week, that would be 156 a year and more than 1%
per year. I never claimed that 1% would die in a year. I claimed that around 1% will die flying i.e. over the course of their flying careers. You're trying to calculate the incidence of death-by-aviation amongst private pilots, which admittedly would be a more sensible thing to do in this instance.
The truth of the matter is that I don't remember where I first found/derived the 1% statistic. I would agree that it's suspect in that there's probably not enough information in the public domain to get more than a decent approximation. However, it doesn't seem outlandish to me. The statement could also be more precisely specified - does '1% of ppls' mean 1% of people who qualify over a given period of time - which will include a lot of inexperienced pilots - or does it mean 1% of people who hold a ppl at a given point in time - which will contain a higher proportion of long-term pilots who have built up a significant number of hours.
Have a look at CAP667, page 30. 158 pilots were involved in fatal accidents over a 10 year period, at the end of which there were 25592 pilots. That's 0.6% fatal accidents per pilot. Of course some of the people killed will not have been the pilots. On the other hand, more than one person is killed in an average fatal crash, and that '+1' person is reasonably likely to be a pilot. I am well aware of the fact that the number of pilots has waxed and waned over the years. I'm also aware that many pilots won't fly for the 16 years necessary to rack up a 1% risk. On the other hand, many will fly for much longer.
Another way of looking at it - aircraft are involved in fatal accidents somewhere between 1/10,000 hours for permit types and 1/43,000 hours for certified types (page 34). 100 hours in a permit aircraft - 1% of 10,000 - isn't that difficult to accrue. Fewer people will get to the 430 hours required to run a 1% risk of being involved in a fatal accident in a certified aircraft. That said, the pilots in the 37 year-old+ age categories were racking up 34-49 hours a year so it isn't unreasonable to imagine that a high proportion of these long-term PPLs would reach or exceed that numbers of hours over the course of their flying careers ('career' meant in the non-commercial sense).
So, as a citeable statistic, mea culpa, I'll try and remember where it came from but I can't quite recall that at the moment. As a decent back-of-the-envelope figure I stand by it (which is what 1/100 is, as oppose to 1/89 or 1/122).
What I would fear is that if an AME has enough pilots on his or her books to make being an AME worthwhile, the chances are that a few of them will be killed in flying accidents over the course of the AMEs career. I think my AME saw about 170 pilots a year - I don't know if that counts as a lot or a little.
Flybymike: the question your statements relate to is 'how many accidents would have been prevented by the requirement to get a medical from an AME?'. If the answer is 'hardly any' then you can make a good argument against requiring a medical. It's an argument I have some sympathy for, with the caveat that I don't know enough to hold any firm opinions in either direction.
'What proportion of accidents are partially or wholly due to medical causes?' is a different question. This paper suggested about 4.7%:
Role of medical factors in 1000 fatal aviation accidents: case note study BMJ 1997;314:1592.
As a prospective AME, the question to ask would be different again: 'if a pilot has a fatal accident, how much grief will the investigation cause me, even if it's ultimately not found to be due to medical causes?' As an example, only a small proportion of fatal accidents are generally thought to be suicide, but what proportion of all accidents
might have been due to suicide? Will my notes be medicolegally watertight? Or if there's a discrepancy between the GP records (e.g. early signs of heart failure) and the AME records (no early signs of heart failure) will this result in aggressive cross-examining whether or not it's directly relevant? Pilots and their passengers are often wealthy and I'd expect a reasonable chance of high-grade acrimony after untimely deaths. I have no idea whether or not this is frequently the case.