Religion as evidence

[Eugene Volokh]

Art Spitzer writes:

> Given the enormous disparity between formal religious doctrine and the
> religious practices of most Americans, it seems unfair to draw much of an
> inference just from the identity of a person's *nominal* religious
> affiliation.

    That seems plausible, but is the claim that it's impermissible to
draw *much of an* inference, or that it's impermissible to draw *any*
inference?  Say the ALJ says:  "I realize that there's an enormous
disparity between formal religious doctrine and religious practices
of most Americans; I therefore realize that the inference from Ms.
Smythe's religious affiliation to her likely mental state is weak.
Nonetheless, it does have some probative value, and though without
this evidence I'd have probably ruled for Ms. Smythe, with the
evidence I rule for the Fair Employment and Housing Commission."
Constitutional?

---------------------------------------------------------------------
"What makes a samurai                         Eugene Volokh
 step down from his high horse?               UCLA Law School
 A few bad quarters."                         405 Hilgard Ave.
 W. Warriner, 101 Corporate Haiku no. 94      L.A., CA 90095