Information, expectations and economic planning

Abstract

Information - understood as items of and in particular accretions to knowledge which act over subjective probabilities attached by economic agents to uncertain events - is a commodity whose production, exchange and use presents special problems of measurement, private and social evaluation of its effects, classification. Treatment of these problems in recent literature on information economics is recalled in order to discuss the formation of expectations - precisely as subjective probabilities modified by information - on the basis of which necessarily economic agents operate in any sequential economy, i.e. an economy where markets - even if generalised to include markets for future and contingent commodities - open and shut repeatedly or are continuously open, making impossible a general equilibrium allocation based on a single price system at a point in time. The informational efficiency of markets, postulated by the theory of so-called "rational" expectations - which would be better defined as "successful" instead - is rejected because i) it is contradicted by the paradox of worthlessness of information which can be derived from it; ii) it implies innate knowedge instead of learning processes; iii) it neglects the possible self-fulfilment of expectations. The informational inadequacy of markets remains at the root of inefficiency, disequilibrium and instability which visibly characterises market economies. Attempts at ex-ante coordination of decisions of economic agents have been attempted with both French-type indicative planning and Soviet-type command planning. Both however come up against not only practical difficulties - which might be reduced by technical progress in information processing and communication - but also the non-cooperative strategies of participants in the planning exercise, though for different reasons and with different implications in the two systems. The alternative remains of delegating to markets detailed decisions about inputs and outputs, limiting policy to macroeconomic variables; this however requires the development of new instruments of control and information (above all about the costs and benefits of such control).