In a message dated Mon, 25 Sep 2000 2:57:08 AM Eastern Daylight Time, Daniel Davies <d_squared_2002 at yahoo.co.uk> writes:
<<
--- JKSCHW at aol.com wrote: > In a message dated 9/23/00
11:19:20 AM Eastern
> Daylight Time,
> jmhayes at j-o-r-d-a-n.com writes:
>
> << Am I the only
> one who thinks that this was written to "prove"
> that the death
> penalty doesn't have a deterrent effect? As if you
> could set up
> a hypothesis that goes "Say it had a deterrent
> effect; then, in
> the places where they have it, you should see lower
> rates of
> homicide. If you don't see that, there's no
> deterrent!"
>
> What lousy logic! >>
>
> Um. Why exactly is the logic lousy? On the surface,
> it looks like Mill's
> methods od co-variation and difference, fundamental
> to science
Looks like, but ain't. Jordan's completely right here (I have gone into one on this subject on this lst a few times, most frequently on more guns and less crime, most recently on Jim H's carbon-monoxide-is-good-for-you numbers).
You can't talk about covariation and difference unless you have an exogenous variable explaining an endogenous one. When you can't rule out the hypothesis that death penalty laws are passed in response to a high murder rate, rather than vice versa, then you don't even know what the model you're testing is, and it's utterly irresponsible to pretend that the numbers you're using say anything interesting.
dd
____________________________________________________________ Do You Yahoo!? Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk or your free @yahoo.ie address at http://mail.yahoo.ie
>>