… is from page 166 of Stephen Marglin’s 2008 book,
The Dismal Science; this book is deeply flawed, in many ways, but Marglin gets this Hayekian point right and he explains it well (the words that Marglin italicizes are a quotation from F.A. Hayek’s 1945 article “
The Use of Knowledge in Society“; the first set of bracketed words are supplied by Marglin; the other sets are supplied by me; link added):
For if knowledge is algorithmic in character, there is no reason for knowledge to become available only if the agent is an active participant in implementing his (or her) knowledge. It is when the knowledge in question has an irreducibly experimental component that
use can be made [of it]
only if the decisions depending on it are left to him or are made with his active cooperation. In today’s jargon, agents have private information, or, rather, hyperprivate information: the knowledge is so private that the knower herself becomes knowledgeable only when translating the knowledge into action.
If people really could formulate all their knowledge in algorithmic terms and calculate as [modern mainstream] economic theory assumes, there would be no need for real-life markets.
[Oskar] Lange‘s pretend markets would do just fine. The virtue of the real market is precisely that it calls forth knowledge that people cannot explain, justify, or defend intellectually, knowledge that economic agents themselves may not fully understand.
It calls forth this knowledge by the incentives it it provides for action and the ruthlessness with which it weeds out error.