[FoRK] Top 5 charitable opportunities

J. Andrew Rogers andrew at ceruleansystems.com
Sun Apr 24 23:51:30 PDT 2011


On Apr 24, 2011, at 10:44 PM, Gordon Irlam wrote:
>> I was pretty sure this number was far too high and likely constructed from an obviously improper set of assumptions but then I went to the referenced sources.
> 
> I withdraw my claim that the $3m number came from Wikipedia.  I think
> I must have been confused.
> 
> Have a look at "The Value of a Statistical Life"
>    http://www.springerlink.com/content/hvx240t1j16vw60w/
>    http://lsr.nellco.org/cgi/viewcontent.cgi?article=1180&context=harvard_olin
> (free)
> 
> Look at tables 2 and 3.  They present 40 estimates of the value of a
> life made by economists.  The vast majority of which lie in the range
> $1-10m, and the geometric mean of which is $3m.


What they are discussing and how you are interpreting it are two very different things -- did you read it? The above reference lends no material support to your assertion.

Are you dissatisfied with the much lower values that come from the previously touted references that you apparently did not read?


> Does that satisfy you that my use of a $2m number for the value of
> life is reasonable, if not conservative?


No. There is no support for it in a casual search of literature, and even if there was the math doesn't work. Constructing a consistent economic picture from this assumption is...difficult. 


> Average that $8-12 cost out to $10.   Now take the $2m figure for the
> value of life.  Divide it by $10, and you get a 1:200,000 value
> multiplier as promised.


Even if I accepted the $2m figure, you still won't arrive at the correct valuation by dividing that figure by the cost of a pill. The arithmetic does not work that way.

That aside, there should be mountains of empirical data to support this multiplier effect in this particular case but you are not using it, which raises its own questions. Given the claimed multiplier, it would be very hard to hide the results.


More information about the FoRK mailing list