The Models and Methodology of Forecasting

Article excerpt

Frequently in the news one hears reports of this or that bank or federal or international agency reducing their economic forecasts. Typically in these revised forecasts the nation is predicted to grow at a still positive, although much reduced, rate.

At the University of Oklahoma Center for Economic and Management Research, we closely follow the forecasts from the Fair Model, created by Yale economist Ray C. Fair and available free of charge at the Web site fairmodel.econ.yale.edu

These forecasts are used as inputs to our indicator-based approach. The real gross domestic product historical data and forecasts, as shown in the GDP table, reveal just how vibrant growth has been since 1995 and extent of the expected decline in that growth rate in the forecast period. Note that the annualized percent change in real GDP has been has high as 6.1 percent since 1995. For 10 of the most recent quarters, growth has exceeded rates that Chairman Alan Greenspan and the Federal Reserve Board of Governors view as unsustainably high without inducing increases in inflation. The Fed generally regards any rate greater than 2 percent as unsustainably high. Fair's forecast shows that the Fed's concerns for too high growth are soon to be alleviated. His results have the U.S. economy growing at only a 0.5 percent rate in the quarter immediately ahead. For 1999, growth by his model is expected to average only about 0.8 percent. It will take until the year 2001 to return to respectable growth rates. To be sure, Fair's results are on the low side of most forecasts for 1999 and beyond. But it is significant that Fair decreased his forecast for 1999 by almost a full percentage point from his prior forecast, just three months ago. If the Fair Model forecasts materialize, there will be repercussions on the Oklahoma economy. We have noted in our publications the greater diversity of growth and the renewed dependency of the Oklahoma economy on continuing national economy growth for its well being. Problems in agriculture and energy markets accentuate the reliance of the Oklahoma economy on the nation for its growth impulses. For Oklahoma to keep growing, we need the nation to keep growing. Behind the indicators The focus of this article will be somewhat different from the typical spouting of economic statistics and forecasts, although such diversions can scarcely be avoided. The hope is to convey the motivation and rationale that underscore the Price College Indicators. These indicators debuted in the 70th anniversary publication of the Oklahoma Business Bulletin in January 1998. The indicators form the core of a forecasting strategy that we use to make economic forecasts. Business cycle research and economic indicators have a long history in economic research. Analysts in this research domain include many famous scholars, Nobel prize winners and former chairs of the Federal Reserve. Examples include works by Burns, Hall, Hansen, Hildebrand, Kindleberger, Lucas, Mitchell, Moore, Schumpeter, Tinbergen, Volcher and Zarnowitz. These are hardly household names to the average citizen, but students of economics know them well. Interestingly, Arthur Barto Adams, founding dean of the OU College of Business Administration and initial publisher of the Oklahoma Business Bulletin, was an early author on the subject. His book Economics of Business Cycles was published in 1925 by McGraw-Hill. He remained active in this area over the years. In 1950 he published Business Cycles: Their Causes and Control. Our current research, then, rests on a long tradition at OU. There are several rationale for undertaking construction of a new series of economic indicators. The financial press, of course, devotes significant attention to every morsel of economic news. The stock market and interest rates rise and fall on each new announcement of housing starts, consumer sentiment, retail sales, gross domestic product, trade deficits, growth in jobs and unemployment rates. …