Risk Averse

Good article in the Atlantic this month (SOTU issue) in which Cullen Murphy ponders David Ropiek's book about Risk assessment (Ropiek and Gray are the Harvard analysts profiled in the Boston Globe recently, the book is Risk: A practical guide for deciding what's really safe and blah blah blah)

Murphy is not used to scientific thought, I would venture -- otherwise she would have seen the absurdity of the "hell-gorithm." And she doesn't seem to grasp the difference between correlation and causation, or how things that correlate in one population may not correlate in another.

I have a feeling Ropiek's book is going to talk about Quality Life Years, and the cost of each, and maximizing the number of QLY (I forget the proper acronym) per dollar. One example I read about (forget where) talks about billions of dollars needed to remove the last trace of arsenic from water, which would only save a few lives and hurt quality in a few others. Hence, the ratio QLY saved or generated, to dollars spent, is comically low. In contrast, seatbelts or child-proof locks or lipitor or fluoride have excellent ratios.

This sounds like a good way to allocate funding and policy priorities. Yet as Easterbrook notes, the media continue to harp on plane crashes and murders (one of his many stats I plan to catologue is that more serial killings occur in hollywood movies than have occured in real life). Yeah, Hollywood is not the media, but you get the point (and Easterbrook has more stats).

But what I hope Ropiek addresses -- and if he doesn't, I have a column -- is that harping on staircase and window accidents may make sense in terms of reducing ER visits, but danger-proofing such everyday items runs so counter to 'our way of life' that it jeapordizes the whole public perception of risk-guided policy decisions.

What I mean to say is -- if the end product of risk-analysis is to have people shun cars, eat wheat germ, avoid sex, and shave every day (there's a correlation between not shaving and getting a stroke) -- well, then people aren't going to pay attention to risk analysis.

We should use risk analysis to make incremental improvements, or, when an opportunity is at hand, to help guide either-or decisions... Otherwise you end up like Ralph Nader, who, if I recall the story correctly, blamed a senator for his wife's traffic death because the senator questioned airbag legislation.

People can tell when you're passionate about saving lives vs. passionate about messing with them. Nader and his ilk aren't so much anti 'personal misery', the pain of burying loved ones or suffering alongside them -- they're often pro-'corporate misery'. Another angle: Naderites have either studied the actuary tables for so long that they decided revolution was necessary to save people, or they had preconceived notions about what people should do, and find data to support it.

In math terms (because I can't get my head around this): we're not just talking about allocating dollars to maximize some $/QLY ratio, we're talking about minimizing changes to some kind of national Quality measurement. Q can't just be for one person's Q, but must factor in the Q of everyone around him. Otherwise, ripping up your street and dismantling your and putting you on lipitor might increase your Q, for very little money, but would hurt everyone else...