Expectations and knowledge assumptions in some popular macroeconomic models

Abstract

Many of the current and past debates in macroeconomics have resulted from an improper understanding of the role of expectations and knowledge assumptions in competing models. In recent research, the author has examined these assumptions in some popular models that have been built during the evolution of macroeconomics. Macroeconomic equilibrium requires two conditions: markets must clear and expectations must be fulfilled. Thus, one requirement for equilibrium is that agents' expectations or "subjective facts" must correspond to the "objective facts."Achieving equilibrium can then be seen partly as a matter of bringing the expectations of the agents into closer and closer agreement with the objective facts. Once an equilibrium is established, it can be maintained only if the agents' expectations continue to correspond to the objective facts. This is the equilibrium definition of Harrod, Hayek, and Lindahl, in which events must be proceeding pretty much as agents expected them to. The author has set up a classification scheme which contains six categories. There are three states of expectation, each of which is subdivided according to market clearing or nonmarket clearing. The three states of expectation are: (i) all agents have the same expectations and they are all correct; (ii) all agents have the same expectations but they are all incorrect; and (iii) agents have different expectations, some of which must obviously be wrong. Full Information lVtacroeconomics corresponds to the case where all agents have the same expectations and they are all correct, and markets clear. All the models surveyed eventually reach this state. WNen this expectations assumption is combined with non-market clearing one has the Neo-Keynesian Disequilibrium models, such as those surveyed by Malinvaud, and, more recently, by Benassy. When all agents have the same expectations but they are all incorrect, and markets clear, one has the New Classical Macroeconomics of Lucas. When this expectations assumption is combined with non-market clearing one has the Neoclassical Synthesis, represented by models such as those by Klein and Patinkin. When the agents' expectations are in disagreement and markets clear one obtains the models of Wicksell, the early Keynes (Treatise), and Friedman. When this expectations assumption is combined with non-market clearing one gets the model of Keynes's General Theory. An example may clarify the nature of the work done. In the Wicksellian pure credit model markets are clearing but agents have different expectations concerning the "natural" rate of interest. In the typical inflationary case studied by Wicksetl the entrepreneurs believe the natural rate to be higher than the banks do. The entrepreneurs borrow from the banks~ thus expanding the money supply, and the cumulative process begins. Wicksell's basic presumption was that the entrepreneurs are usually right and the banks wrong. Thus, the resulting inflation will come to an end only when the banks revise their expectations and bring them into line with the objective facts.