Note on Starr's analysis.
I would have said, for any economic activity:

value = expected return = monetary cost + adjusted value of incidental risk.
a b c . d

He seems to say that

d~ ¢ which, as a general proposition, has no evident logical basis.
BUT

He probably is right to deny that d>>c in any situation where 1) d is allowed
to fluctuate without evident impact onc,

or 2) where the
participants aver that d is negligible compared to c, which might be implied
by finding that demand is quite elastic in response to c.

There remain the problems of

1) information and rational perception of d
2) thrill values -- linked to d
3) proper bounds for a defined activity over a domain where c and d may
fluctuate widely (auto or gun safety, for example).
or knives.

(Do we lump all knives, or classify them by length and sharpness
of blade.)

Prof. Joshua Lederberg
Department of Genetes
School of Medicina
Stanford University
cuntad, California 94305