Today's quote is from Martin Landau's classic 1969 essay "Redundancy, Rationality, and the Problem of Duplication and Overlap."
"In complex and tightly ordered systems, the
cost of error can run very high...
“Is it possible to take a set of individually
unreliable units and form them into a system “with any arbitrarily high
reliability”? Can we, in other words, build an organization that is more
reliable than any of its parts?
“The answer, mirabile dictu, is yes. In what is now a truly classical paper, Von
Neumann demonstrated that it could be done by adding sufficient redundancy.”
Landau was way ahead of his time on applying complexity theories to organizations.
Bonus quote today is found on page 510 of Philosophy of Complex Systems, from an essay by John Foster called "Economic Systems" (2011):
"Over the past half
century, economic theorists have stubbornly held on to their view that
economics should be a branch of decision theory that involves optimizing
choices along artificially smooth and conveniently specified production and
utility functions, subject to constraints."
A lot was said in that sentence. Economists were "stubborn" because, despite the continual failure of their models to describe reality, they hold on to assumptions of "artificially smooth" and "conveniently specified" functions. They hold onto these methods because linearity is mathematically tractable (the same critique is sometimes levied on physics).
But clearly we live in a complex world dominated by numerous changing factors. For this reason, it is misleading to think of economics as "a branch of decision theory that involves optimizing choices" and instead one should think about how individuals interact and exchange.
As Armen Alchian concluded in 1954:
"Systems analyses are machines for generating
implications of postulated initial information; they do not generate decisions... Under uncertainty, the criterion of decisions is
not simple maximizations; the essence of the decision process is to affect the
scope of random factors so as to give a “good” probability distribution of
outcomes. The insurance principle is to decisions what maximizations are to
analytic implications."
His 1950 paper "Uncertainty, Evolution, and Economy Theory" argued that optimal decisions can only be regarded as such after-the-fact through a process of natural selection. Decision problems under uncertainty were solved, he later found, by maximizing the option space available. He agreed with Herbert Simon that one should optimize when one has the necessary information, but an evolutionary "satisficing" method probably more appropriate given the vagaries of our world.
