-
Well, you rely on very complex computer models to do that,
-
so that you have to faith in those who are doing the
-
projection and the quality of the analysis.
-
So that this problem can be kind of capsulized in the idea
-
that costs are often relatively well known,
-
and they can be projected with relative precision over the
-
short term compared to the harms,
-
the environmental and health harms,
-
that are highly uncertain.
-
And they're often projected way out into the future with a
-
really fuzzy image of how they're going to fall out
-
differently on different groups.