Bryan Caplan links to Robin Hanson’s comparison of "the abject deference the public gives to physicists with the stubborn defiance the public gives to economists." Robin is right-on. His comments called to my mind this passage from Chapter I, Part I of Frank Knight‘s great 1921 book Risk, Uncertainty, and Profit:
Finally, it makes vastly more difference practically whether we
disseminate correct ideas among the people at large in the field of
human relations than is the case with mechanical problems. For good or
ill, we are committed to the policy of democratic control in the former
case, and are not likely to resort to it in the latter. As far as
material results are concerned, it is relatively unimportant whether
people generally believe in their hearts that energy can be
manufactured or that a cannon ball will sink part of the way to the
bottom of the ocean and remain suspended, or any other fundamental
misconception. We have here at least established the tradition that
knowledge and training count and have persuaded the ignorant to defer
to the judgment of the informed. In the field of natural science the
masses can and will gladly take and use and construct appliances in
regard to whose scientific basis they are as ignorant as they are
indifferent. It is usually possible to demonstrate such things on a
moderate scale, and literally to knock men down with "results." In the
field of social science, however, fortunately or unfortunately, these
things are not true. Our whole established tradition tends to the view
that "Tom, Dick, and Harry" know as much about it as any "highbrow";
the ignorant will not in general defer to the opinion of the informed,
and in the absence of voluntary deference it is usually impossible to
give an objective demonstration. If our social science is to yield
fruits in an improved quality of human life, it must for the most part
be "sold" to the masses first. The necessity of making its literature
not merely accurate and convincing, but as nearly "fool-proof " as
possible, is therefore manifest.
