3
out of a theory, hut theory out of economic development.
We like to think of theory not as creating experience,
but helping us to understand experience. At best, the
rationalist mind of pure theory is an imperfect medium to
solve a development problem, which is so very much based
4
upon unique historical experience and empirical data.
Similarly, numbers and trends should be viewed
with caution and accepted only for what they are--the
results of underlying phenomena, and not the phenomena
themselves.
. . one might suggest that the statistician needs
to be receptive to the results of the analytical
theorists, to the suggestions of the student of the
historical scene, and even to the claims and clamor
of the reformers. And he must beware especially of
the danger of identifying mechanically derived lines
with trends; calculated ratios with immutable and
natural laws of constitution, and correlation
coefficients with inviolable laws of causation and
association.(5)
Xn recent months all these fears have been re
echoed by leading members of the profession.^ If economic
The contrast is most sharp in the two disci
plines anthropology and economics. The former is still
essentially empirical; the latter has become the most
abstract and mathematical of all the social sciences. If
at the moment there is a certain disillusionment in the
economics profession, perhaps it is because we are
expecting of theory that which it cannot give.
5
Simon Kuznets, Economic Change, Selected Essays
in Business Cycles, National Income, and Economic Growth
(New York: W. Norton & Company, Inc., 1953)> pp. 294-95*
^Cf. Wassily Leontief, "Theoretical Assumptions
and Nonobserved Facts," American Economic Review. LXI
(March, 197l), 1-7; Joim G. Gurley, "The State of
Political Economics," American Economic Review, Papers
and Proceedings of the Eighty-third Annual Meeting of the