Rahul Mahajan asks 7/11 why James Craven objects to libertarians' being
"deductivists":
>All of science is done by taking the basic underpinnings (axioms,
equations of motion, etc.), deducing consequences of them, then
empirically testing those consequences. This is the only way to do it.
The eq.'s of motion themselves cannot be "directly" tested -- they can
only be tested by their results.
>In what imaginable sense is this specific to libertarianism?
Libertarian economists (or libertarians considering economics) share
with their non-libertarian neoclassical colleagues a marked tendency to
regard only quantifiable things as worthy of attention. They do not
pause to wonder whether the question "is economics a science"? or "if
it's science, must it be done in numbers"? or "if it's a science, must
it _all_ be done in numbers"? or "if economics is not a 'science,' does
whatever it is require that everything be done in numbers"? Mahajan's
query assumes that economics is not only a science, but the kind of
science which requires and permits only quantifiable things to matter.
Neoclassicals do it because it makes their theory tidy, and provides
ample opportunities to crunch numbers to generate publishable papers (too
much reading David Colander there). Libertarians do it because once they
concede the importance of "institutions" (in the comprehensive sense
that, say, North uses the term), the atomistic libertarian game is up:
others must be taken into account because in an unquantifiable but
unavoidable sense, each of us includes the others.
Michael Etchison
[opinions mine, not the PUCT's]
