-
So thinking about whether or not
-
it makes sense to give an administrative agency
-
that kind of discretion is very important.
-
And it depends again on the technical nature of the problem.
-
Another key question is risk averaging.
-
If you have lousy data, data that is collected maybe
-
with very small sample sizes so that you have no understanding
-
of demographic variability and exposure or in the complex
-
temporal and spatial distributional patterns of pollution,
-
then it's like looking at the world without my glasses on.
-
Which is, all the students look the same
-
so that you lose any resolution,