I just read a working paper by Joe Kane, ‘The Economic Flaws in Computerized Socialism’, which refers to Eden Medina’s excellent 2011 book about Project Cybersyn in Allende’s Chile, Cybernetic Revolutionaries. She recounts the history of this attempt to use computers for central planning, from a futuristic control room, relying on data input in factories around the country. Even before the coup, the project was in some trouble. As Kane notes, there has been a revival of the idea that computers and the internet (and now the blockchain) make central planning feasible. Evgeny Morozov trailed the idea an article that drew on Medina’s book. So did Paul Mason in his dire book Post-Capitalism. Kane cites a few other examples.
As he points out in the working paper though, the central planning problem is not one of computation or the transfer of information – or rather, that is only the case in the world of neoclassical general equilibrium theory. Frances Spufford’s brilliant book Red Plenty makes use of the formal equivalence of the centrally planned and competitive market economies in that la-la-land with no frictions, fixed preferences and complete markets. Markets are a process of discovery, as John Kay has to repeat over and over again. The data that would be needed for computerized central planning are not ‘out there’, but are created by the market. I wouldn’t go nearly as far as Kane in concluding government policies are therfeore useless, but do agree that “the outcome of the market process is not separable from the process which generates it.”