ECONOMIC FORECASTING

the prediction of any of the elements of economic activity. Such forecasts may be made in great detail or may be very general. In any case, they describe the expected future behaviour of all or part of the economy and help form the basis of planning. Formal economic forecasting is usually based on a specific theory as to how the economy works. Some theories are complicated, and their application requires an elaborate tracing of cause and effect. Others are relatively simple, ascribing most developments in the economy to one or two basic factors. Many economists, for example, believe that changes in the supply of money determine the rate of growth of general business activity. Others assign a central role to investment in new facilitieshousing, industrial plants, highways, and so forth. In the United States, where consumers account for such a large share of economic activity, some economists believe that consumer decisions to invest or save provide the principal clues to the future course of the entire economy. Obviously the theory that a forecaster applies is of critical importance to the forecasting process; it dictates his line of investigation, the statistics he will regard as most important, and many of the techniques he will apply. Although economic theory may determine the general outline of a forecast, judgment also often plays an important role. A forecaster may decide that the circumstances of the moment are unique and that a forecast produced by the usual statistical methods should be modified to take account of special current circumstances. This is particularly necessary when some event outside the usual run of economic activity inevitably has an economic effect. For example, forecasts of 1987 economic activity in the United States were more accurate when the analyst correctly foresaw that the exchange value of the dollar would fall sharply during the year, that consumer spending would slacken, and that interest rates would rise only moderately. None of these conclusions followed from purely economic analysis; they all required judgment as to future decisions. Similarly, an economist may decide to adjust an economic forecast that was made by traditional methods to take account of other unique conditions; he may, for example, decide that consumers will alter their spending patterns because of special circumstances such as rising prices of imports or fear of threatened shortages. Although judgment may be based on experience and understanding, it may also be no more than unconscious bias. Forecasts based on judgment cannot be subjected to the kind of rigorous checks applied to forecasts developed by the use of more objective techniques. Consequently, the most accurate and useful forecasts are likely to be those founded on essentially economic considerations and standard statistical techniques. Though they can then be modified by the application of judgment, the resulting changes should be stated explicitly enough so that anyone wishing to use a forecast will know where, and how, it has been affected by the forecaster's own judgment, or bias. Economic forecasting is probably as old as organized economic activity, but modern forecasting got its impetus from the Great Depression of the 1930s. The effort to understand and correct the worldwide economic disaster led to the development of a vastly greater supply of statistics and also of the techniques needed to analyze them. After World War II, many governments committed themselves to maintaining a high level of employment. Most governments of the industrialized Western countries were prepared to intervene more often and more directly in economic affairs than previously. Business organizations manifested more concern with anticipating the future. Many trade associations now provide forecasts of future trends for their members, and a number of highly successful consulting firms have been formed to provide additional forecasting help for governments and businesses. the prediction of future economic activity and developments. Forecasting of this nature has grown rapidly since the 1930s, largely in response to the increasing unpredictability of the economic situation, a greater involvement of governments in economic affairs, which requires the preparation of economic plans and projections, and rapid improvement in the quality and coverage of economic statistics and forecasting techniques. There is a vast array of forecasts available, ranging from short-term predictions for specific economic variables (such as interest rates) or of demand for individual products (such as steel or automobiles) to medium- and long-term forecasts of the economy as a whole. Despite their lack of certainty, such forecasts are widely used in business, government, and private affairs to help in formulating policies, strategies, legislation, and long-term plans. A useful distinction can be made between macro- and micro-economic forecasts. Macroeconomic forecasts are designed to predict the future course of the entire economy or of specific broad economic variables, whereas microeconomic forecasting is designed to project the likely development of particular economic sectors such as one industry, commodity, or firm. The best known and most widely used form of macroeconomic forecast is that of national income or gross national product (GNP). This predicts, in numerical terms, the major components of a country's economic activityprivate consumption, government expenditure, private and public investment, and the balance of exports and imports. All countries devote significant resources to this type of forecasting, typically on a one- to five-year basis. These forecasts are used for a number of different purposes. Governments use them to determine future economic strategy and to predict other variables of the economy such as the likely level of inflation, industrial output, employment, etc. Based on such a forecast, the effects of various proposed government actions (a cut in taxation or an increase in government expenditure, for example) can be tested before official policy is finalized. Macroeconomic forecasts are also used outside government as a basis for producing more detailed projections of the main components of the economy and in the preparation of microeconomic forecasts. By studying the overall forecast of private consumption, for example, a retailer might, by referring to established patterns of spending, predict the amount that is likely to be spent on foodstuffs and nonfood products and then, working in the microeconomic forecasting area, attempt to determine future expenditures in specific product categories. Similarly, an automobile manufacturer will attempt to predict demand for his product by looking at the predicted level of and trends in disposable incomes and consumption and predictions of interest and exchange rates. He will also forecast production costs from the trend of wage increases and inflation. In general, most microeconomic forecasting starts with some forecast or assumption about the economy as a whole that is then modified or analyzed into its components in light of special factors and considerations applicable to a particular product, industry, or other concern. Forecasts range from one month to 10 years or more. However, largely owing to the economic shocks of the last 20 years (e.g., the quadrupling of oil prices in 1972), there has been a trend away from highly numerical long-term forecasts in favour of indications of the broad direction of economic developments, based on both statistical evidence and more or less subjective judgments on such basic aspects of the economy as population growth, technological progress, and social changes. A set of long-term forecasts may be made to indicate the likely outcomes of several different but equally plausible assumptions in a technique often called scenario building. The techniques of economic forecasting have developed rapidly in recent decades. This in part reflects the growing understanding of the ways in which numerous economic variables affect each other. Equally important reasons are the better availability of good statistics and the development of computer methods for processing large amounts of data. Computer capacity has made possible the practical development of mathematical models of the economy through which it is possible to explore the relationships between the key determinants of the economic system with a speed and to a degree of detail that were not possible before. Most governments and large forecasting organizations use computers in this technique known as econometric forecasting. Additional reading Various of the economic indicators commonly used for forecasting are surveyed in Kenneth C. Land and Stephen H. Schneider (eds.), Forecasting in the Social and Natural Sciences (1987); John Llewellyn, Stephen Potter, and Lee Samuelson, Economic Forecasting and Policythe International Dimension (1985); and Geoffrey H. Moore, Business Cycles, Inflation, and Forecasting, 2nd ed. (1983). Techniques for selecting appropriate data and models for use in economic forecasting are discussed in Giles Keating, The Production and Use of Economic Forecasts (1985); Lawrence R. Klein and Richard M. Young, An Introduction to Econometric Forecasting and Forecasting Models (1980); Steven C. Wheelwright and Spyros Makridakis, Forecasting Methods for Management, 4th ed. (1985); and Norman Frumkin, Tracking America's Economy (1987). One of the most useful sources of timely statistical data for U.S. economic forecasts is Survey of Current Business (monthly). Richard W. Everett The Editors of the Encyclopdia Britannica