Expectations, Implicit

BIBLIOGRAPHY

The concept of implicit expectations, which was pioneered in 1954–1955 by Edwin S. Mills, circumvents a problem encountered by economists who study the behavior of business enterprises: Although production and pricing decisions are made on the basis of expectations about future economic conditions, economic researchers must cope with a shortage of observations on actual sales anticipations. One research strategy is to work with explicit assumptions about the way expectations are formed. Thus, in 1953 Robert Ferber concluded on the basis of his examination of empirical evidence in the Railroad Shippers’ Forecasts that expected sales are equal to actual sales of the same period of the preceding year adjusted for recent trend. In contrast, Mills did not attempt to model the process by which expectations are generated. His implicit expectations approach substitutes the actual realized sales—what sales turn out to be—for the unobserved expected sales under the assumption that the forecast error is random. That is to say,

E_t –X_t = ε_t,

where E_t is expected sales, X_t is what sales actually turn out to be, and ∊_t is a random error with zero expected value (it averages out to zero). Mills used his implicit expectations concept in a number of theoretical and empirical articles, including a 1957 study in which he was able to estimate the magnitude of undesired inventories generated by sales forecast errors made by American corporations.

Mills’s implicit expectations approach is closely related to John F. Muth’s (1961) concept of rational expectations that Robert Lucas invoked in 1972 in spearheading a revolution in the field of macroeconomics. Muth, like Mills, assumed that forecast errors are random. This is an example of independent discovery, as often is encountered in scientific research (Merton 1973), but a rather subtle difference may distinguish the two concepts. Expectations are implicit if the random forecast error ε_t is distributed independently of E_t but rational if they are distributed independently of X_t. If a firm’s sales anticipations are based on a random survey of customers about their expected demand, the error in forecasting total sales that results from sampling error will be distributed independently of what sales actually turn out to be; this means that the sample survey has generated an implicit forecast of sales. If instead a forecast is generated with data from prior years with a multiple regression of actual sales on observable variables, such as the unemployment rate, price, and sales in prior periods, the forecast will be rational because the errors of a correctly formulated regression forecast, as a mathematical necessity, are distributed independently of the prediction. Implicit forecasts can be converted into rational forecasts with the equation obtained from historical data by regressing the realization on the forecast.

In a paper published in 1985 Muth reported that his empirical investigations indicated that expectations are not rational: The forecast errors were correlated with the forecasts, and the variance of the forecasts was often larger than the variance of the variable that was being predicted. However, Muth found that the errors also were correlated with the actual realizations; that meant that they were not implicit expectations. He advanced as a synthesis an “errors in the variables” model that allowed the forecast error to be correlated with both the variable being forecast and the actual outcome but cautioned that this model was not fully in accord with the facts. Although Muth became skeptical about his own theory, it provides the foundation for macroeconomic studies of inflation and business cycle fluctuations.

SEE ALSO Adaptive Expectations; Expectations; Expectations, Rational; Expectations, Static; Mills, Edwin; Rationality

BIBLIOGRAPHY

Ferber, Robert. 1953. The Railroad Shippers’ Forecasts. Urbana: University of Illinois Press.

Lucas, Robert. 1972. Expectations and the Neutrality of Money. Journal of Economic Theory 4 (2): 103–124.

Merton, Robert K. 1973. The Sociology of Science: Theoretical and Empirical Investigations. Chicago: University of Chicago Press.

Mills, Edwin S. 1954–1955. Expectations, Uncertainty and Inventory Fluctuations. Review of Economic Studies 22 (1): 15–22.

Mills, Edwin S. 1957. Expectations and Undesired Inventory. Management Science 4 (1): 105–109.

Mills, Edwin S. 1962. Price Output and Inventory Policy: A Study of the Economics of the Firm and Industry. New York: Wiley.

Muth, John F. 1961. Rational Expectations and the Theory of Price Movements. Econometrica 29 (3): 315–335.

Muth, John F. 1985. Properties of Some Short–run Business Forecasts. Eastern Economic Journal 11 (3): 200–210.

Young, Warren, and William A. Darity Jr. 2001. The Early History of Rational and Implicit Expectations. History of Political Economy 33 (4): 773–813.

Young, Warren, Robert Leeson, and William A. Darity Jr. 2004. Economics, Economists, and Expectations: Microfoundations to Macroapplications. London and New York: Routledge.

Michael C. Lovell