Topics in Disequilibrium Economics

This article describes how economic agents rationally formulate their demands when they perceive signals of quantitative disequilibrium. A general definition of effective demand is given for all types of rationing schemes. It is shown to include previously used definitions when rationing is nonmanipulatable, i.e. when each agent faces fixed bounds on his trades. If, on the contrary, he can manipulate these bounds through his demands, rational behavior is shown to lead to an explosive phenomenon of overbidding. Finally the theory is extended to include uncertainty and transaction costs.

Recent works on the microeconomic foundations of macroeconomics (Barro & Grossman, Benassy, Malinvaud, Younès) make use of the Hicksian fix-price method. In these studies, prices are assumed to be temporarily fixed in the short run, and adjustments take place through quantity rationing. A basic assumption of these models is that only agents on the “long side” of a market are rationed (e.g. buyers when there is an excess demand). This paper surveys some recent contributions to the study of the logic of this assumption. It is argued in particular that, although these models are designed to describe “disequilibrium” states where there is an excess demand or supply on some markets, their logic appears to involve some kind of recontracting process at fixed prices among traders.

This paper analyzes the allocation of risk which is associated with variations in the value of aggregate output in a context of differences in worker and firm attitudes to risk. The analysis emphasizes two factors—limited firm risk-absorbing capacity and worker unreliability—which make it infeasible for risk-neutral firms to relieve risk-averse workers of all risk. The main ideas are that the limitation on current wage payments imposed by the value of current output implies that firms generally cannot guarantee their workers a stable consumption stream and that firms generally earn rents in their risk-absorbing role. In addition, differences among workers in their reputation for reliability, which relate to their behavior when the value of output is high, produce differences in the terms at which they can obtain consumption when the value of output is low. Specifically, workers of unproven reliability obtain less consumption in bad states of nature and, unless they plan to be unreliable, have lower current expected utility than workers of proven reliability. Those workers who earn reputations for reliability tend to be those who are relatively more risk averse and those who belong to groups which have a relatively good record for reliability.

In this paper one considers an economy in which individuals can transact at “false” prices. When they do they encounter quantity constraints, this in turn, as Arrow has noted, stops them acting as perfect competitors. In particular they must form an hypothesis concerning a possibility of affecting their quantity constraints by a change of price. This hypothesis is called a conjecture. A set of prices and quantity signals at which desired trades are achieved and no prices change is advantageous under the conjecture is a conjectural equilibrium. Economies can have conjectural equilibria even when they have a Walrasian one. Naturally one wants a theory of conjecture formation. In what follows it is shown, mainly by examples, that it is not fruitful to look for rational conjectural equilibria (defined in the sequel). One concludes that at best one could hope that agents conjecture Marshallian schedules.

The Barro-Grossman theory of the demand and the supply multiplier is generalized in the following ways: Household behavior under disequilibrium is derived from the knowledge of the behavior under equilibrium using the Tobin-Houthakker theory of rationing. The setting of prices and wages under equilibrium and disequilibrium conditions is analyzed. It is shown that the value of the demand multiplier depends on whether employment is instantaneously adjusted during a depression. The results of the microeconomic analysis are used for constructing a macroeconomic model where the interactions between inflation and quantitative adjustments are studied.

A disequilibrium version of the standard IS-LM model is constructed and used to analyze the stability of the model. The main feature of the disequilibrium behavior turns out to be a spillover from the money market to the output market; if there is an excess demand for loanable funds (or an excess supply of bonds) firms will be unable to finance all of their desired investment, and aggregate demand will therefore be smaller than would otherwise be the case. This spillover has certain implications for the dynamic behavior of the model.