"For a successful technology, reality must take precedence over public relations, for nature cannot be fooled."
This is the last sentence of Richard Feynemans excellent dissenting report into the Space Shuttle Challenger Disaster.
I respectfully suggest that this report is worth reading in the context of the NAS system.
http://www.ralentz.com/old/space/feynman-report.html
Richard makes the point that the use of statistics by managers can cause error - because managers and technologists can be caught by the logical error that just because something has not occurred YET, then as time goes on it is less likely to occur.
In the case of Challenger, it was the O-Ring seals in the solid rocket boosters that failed, causing hot gas to escape and ultimately failure of the booster and spacecraft. These seals had eroded on previous missions, but had never actually eroded enough to cause failure until the Challenger Disaster. The mistake that NASA managers made was to assume that because complete failure had not occurred in previous missions, then it would never occur.
I think this is worth remembering in the context of the NAS debate by all parties connected with it.
I have skimmed Prof. O'Niells report. I have not read the NAS risk assessment document, but I am sufficiently disturbed at the methodologies apparently used by BOTH sides in this debate to post a hopefully helpful link to Richard Feynemans brilliant observations on the use and misuse of statistics.
To put it another way, I think safety is too important to be a political football. I do not wish to wake up one morning to very bad news, followed by a public enquiry that establishes that the risks involved in NAS were understated.
In addition, in my past professional life risk managment was a tool that is used to compute the cost and probability of accidents as a means of calculating the amount that should be invested in preventing said accidents. I am not sure what is today decided to be an "acceptable" risk, but I would believe it used to be in the order of 4,000,000 to one.
I therefore ask if the monetary consequences of a major collision are outwieghed by the savings that may occur under a changed system, on a discounted cash flow basis?