Peter wrote:
If true, there was clearly a progression towards the $100k figure, and all GA mfg activity would have stopped long before it got to 100k because any business like that becomes a complete farce. Most planes in question were selling for way below 220k in 1988. For example I know that a C172 was offered by CSE in the UK for £120k to a school at Shoreham, in 2003, which is a very long time after 1988. So the 100k/1988 claim is clearly complete bunk.Looking at this from a different angle, such astronomical costs would imply that probably every plane made had been crashed and the mfg sued.
So I think that wiki article is just another circular reference, typed up by somebody with an interest, and full of bollocks (like a good % of wiki articles, and the internet generally).
I think the progression happened quite fast within a short time. And Cessna did stop SEP mfg activity in the early 80’s. They only restarted mfg with the passing of new liability laws. My memory is that there were around 7-10’000 Cessnas being sold per year at the end of the 70’s. IIRC, one cause was a few cases that made Cessna liable for retroactively applying safety changes/improvements in later models to all aircraft built since the start of production 40 years earlier. This made their insurance costs skyrocket at the beginning of the 80’s and they needed to recover that from new aircraft sold, which of course dropped dramatically as the price soared. So they stopped production.
Fact is that product liability issues killed the US airplane industry. Even the GA revitalization act could not reverse the damage.
The figures are well documented but the results described are speaking an even clearer language.
And when Gara finally became law, in the mean time the certification costs had gone through the roof per airframe as they needed to be recovered with much less airframes.
The final result is still obvious. 1m Sep airframes in tens of numbers instead of hundreds.
At work we pay for product liability insurance annually in advance, according to the sales numbers in the past year.
So if e.g. you were selling $100M, and paying say $1M (1%) insurance, and your sales drop to $1M, then yes you are “paying” as much for product liability as you are getting for each plane you sell. But obviously that is a completely disingenuous way of putting it – nothing short of complete and total deliberately misleading bollox 
The following year you will tell your insurer your sales were $1M and they will drop the premium to $10k (1%). Talking of planes, $1M might be 20 planes (at the time) which is probably about right, when GA sales fell off the cliff! Piper have sold 20 planes a year in quite a few years
It’s obviously nonsense and probably in several ways at the same time.
On reflection – this is even worse than the statistic in “the killing zone”, which should be used in statistics 101 as a textbook example of not adjusting for the sample.
The 100k is also in the same report. But the way these numbers are presented is very poor, even if they are true. In fact, whoever wrote the 50 —> 100k increase, implying like-for-like, needs to have their statistics license withdrawn for life, assuming they had one in the first place.
The wording implies a 2,000 fold increase in insurance cost. Adjusted for inflation (50$ in 1962 is around $200 in 1988) it is “only” a factor of 500.
Then we need to adjust for what constitutes a “general aviation aircraft”, which here includes business jets. Assumes that the definition in the study is similar to the one used in the Wikipedia article, the unit price of an aircraft went from 100-200k in 1962 to 3 million in 1988 (in 2009 dollars), and 100k 1988 dollars are around 180k in 2009. That means 6% of the aircraft price for insurance – quite a bit, but not entirely as bad.
This compares to 240 2009$ for an average $100-200k per aircraft, which is 0.1-0.2% So the actual increase was around a factor of 30-60, not 500.
Not adjusting for the sample, and not adjusting for inflation, are both sackable offences for anyone dealing with statistics.
But there may even be another big issue. I don’t believe product liability insurance is paid up front per aircraft when it is sold [I may be wrong here], so what is much more likely is that the “$100k per aircraft” was the total premium paid in a year divided by the number of aircraft delivered. Given the market crashed from a peak of around 17,000 in the late 70s to less than 2,000 by 1988, and the companies were still paying premiums to insure the liability for the legacy fleet, of course the cost per aircraft sold went up. If they sold no aircraft, it would have been infinite.
Corrected for that (if indeed this is how premiums are paid) the actual increase factor was considerably less, definitely below 10.
So one way or the other, whoever wrote that particular nugget of statistics was wrong by a factor of at least 20, possibly 200.
I think you need to read what I wrote, specifically the bit about the 100k / 1988. The $50 is of course reasonable.
Peter wrote:
So I think that wiki article is just another circular reference, typed up by somebody with an interest, and full of bollocks (like a good % of wiki articles, and the internet generally).
How can you possibly claim that a US government report from 1993 is “just another circular reference”!? With two clicks in obvious places on the Wikpedia page you can have the report in your web browser. The Wikipedia claim you quoted is essentially itselfa quote from the report. (Well, the report does say “roughly $51” and not “approximately $50”.)
Of course, you don’t have to believe what’s in the report…
I simply don’t believe a lot of the stuff in that wiki link e.g.
Average cost of manufacturer’s liability insurance for each airplane manufactured in the U.S. had risen from approximately $50 per plane in 1962 to $100,000 per plane in 1988, according to a report cited by the Bureau of Labor Statistics, a 2,000-fold increase in 24 years
If true, there was clearly a progression towards the $100k figure, and all GA mfg activity would have stopped long before it got to 100k because any business like that becomes a complete farce. Most planes in question were selling for way below 220k in 1988. For example I know that a C172 was offered by CSE in the UK for £120k to a school at Shoreham, in 2003, which is a very long time after 1988. So the 100k/1988 claim is clearly complete bunk.
Looking at this from a different angle, such astronomical costs would imply that probably every plane made had been crashed and the mfg sued.
So I think that wiki article is just another circular reference, typed up by somebody with an interest, and full of bollocks (like a good % of wiki articles, and the internet generally).
For more info on the US liability issue and the associated legal action taken in the 1990s, may I direct everyone to GARA: https://en.wikipedia.org/wiki/General_Aviation_Revitalization_Act
Agreed, but (this has come up many times, in various fora) nobody ever found significant insurance premiums either.
I vaguely recall somebody finding Garmin paying some low single digit percentage of their revenue on an item called “insurance”, which (given that this would include all of their business insurance cover) would be about right. Obviously somebody with time on their hands could check this fairly easily… Garmin is a big company.
In my business we pay about 1% of sales, which is probably same order of magnitude as most companies who actually make something useful.
Like a lot of things, we have done this one already
